The desktop and the Hermes backend it talks to version independently — a
remote VM running an older build still serves display_metadata as JSON text.
Indexing into that string with `in` threw and failed the whole resume, so
narrow the type to admit a string and parse it before reading task_count.
Falling back to the generic label keeps a delegation event renderable even
when the metadata is unusable.
Co-authored-by: xxxigm <xxxigm@users.noreply.github.com>
Co-authored-by: Studio729 <Studio729@users.noreply.github.com>
export_session() reads through get_messages(), so before the read fix an
already-serialized string went straight back into _insert_message_rows() and
got re-dumped — an export/import round trip permanently corrupted the row.
Guard the three write paths the same way tool_calls already is: parse a
string argument before storing it, and drop metadata that isn't an object
rather than persisting something no reader can use.
Co-authored-by: xxxigm <xxxigm@users.noreply.github.com>
Co-authored-by: aml1973 <aml1973@users.noreply.github.com>
get_messages(), get_messages_around() and get_anchored_view() returned the
raw display_metadata column instead of the dict every caller expects. The
desktop paints a resumed transcript from the REST prefetch, which reads
through get_messages(), so any session holding an async_delegation_complete
event failed resume with "Cannot use 'in' operator to search for
'task_count'" — on every such session, not just corrupted ones.
Route all four read paths through one shared codec that also unwraps rows
carrying a second JSON layer, so sessions already broken on disk recover on
read rather than needing a migration.
Co-authored-by: Studio729 <Studio729@users.noreply.github.com>
Co-authored-by: aml1973 <aml1973@users.noreply.github.com>
Co-authored-by: xxxigm <xxxigm@users.noreply.github.com>
Desktop boot polls /api/status, so readiness waits on gateway config and a
cold plugin import tree. On Windows that regularly outlives the probe and
Desktop kills a backend that is already listening, respawns it, and re-pays
the same import cost — the reported crash loop.
Probe /api/health instead, falling back to /api/status only for the
missing-route shapes the fetch helpers emit (404, or HTML from the SPA), so
an older remote backend still connects. Timeouts and server errors keep
polling health rather than dropping to the heavyweight route.
A cheap route is not enough on its own. Warming the gateway import holds the
GIL, so the event loop can stall for tens of seconds and starve /api/health
too. At the default 15s socket timeout only three attempts fit in the 45s
budget; give each probe 5s so the loop keeps retrying across the stall.
Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com>
Co-authored-by: DESXIE <78300229+DESXIE@users.noreply.github.com>
Co-authored-by: frohsinnllc <231045016+frohsinnllc@users.noreply.github.com>
/api/status is the only public liveness route, and its handler loads the
gateway config, probes gateway health, and counts sessions before it can
answer. That work is wrong for a readiness probe: a caller that only needs
to know the process is up pays for a cold plugin import tree.
Add /api/health, which returns process liveness, version, and the auth-gate
shape and touches nothing else.
The create dialog used the settings-surface ListRow/ToggleRow inside a
modal, which read differently from every other form dialog, and the detail
header carried an enabled/disabled pill that rendered as a stray dash.
Switch the form to the shared Field primitive (+ Switch) and remove the
pill.
Blueprints lived behind a separate Jobs/Blueprints tab with its own card
gallery — a bespoke surface no other overlay uses. Remove the tab and make
blueprints a "Start from" dropdown at the top of the New Job dialog
(default "Custom" = the manual editor); picking one swaps the form for that
blueprint's typed slots. Also promote the detail-view "Trigger now" button
to a primary action and adopt the shared Field primitive.
Overlay panes each set their own top padding, so the Settings sidebar and
Panel headers sat at different heights than System/Agents and the close X
(the #67759 regression). Hoist the shared beside-the-X clearance into
OVERLAY_TOP_CLEARANCE, keep the taller pad only on OverlayMain (which sits
under the X), tighten OverlayMain's gutters, and drop the one-off Settings
override. Also give PanelAction a `primary` variant so a detail header can
promote its main action to a filled button.
Dialog forms each hand-rolled their own label+control+hint stack (or
borrowed the settings-surface ListRow), so gaps and hint styling drifted
between the profile, cron, and webhook dialogs. Add a single Field /
FieldHint primitive for label-over-control dialog fields and adopt it in
the create/rename profile dialogs as the first consumers.
The cherry-pick auto-resolution + AST sweep applied plain utf-8 at three
.env reader sites where the salvaged PRs (#62617, #62123) deliberately
use utf-8-sig — a Notepad BOM must not hide/duplicate the first key.
Restore the contract (tests pin it).
AST-driven pass over every Path.read_text()/write_text() without an
explicit encoding= across non-test code: 71 sites in 34 files
(skills_hub, hermes_cli/main+profiles+service_manager+container_boot,
mem0/hindsight/honcho plugins, achievements dashboard, release/CI
scripts, productivity+comfyui skill helpers, agent/*). Verified zero
positional-encoding collisions before insertion; per-file compile()
check after.
Adds a check-windows-footguns rule flagging bare single-line
read_text/write_text (multi-line forms stay covered by the AST guard
test from #38985). Together with the salvaged contributor commits this
retires the ~169-site bare file-I/O class (#37423's long tail).
The bundled office skills (#68595) read user documents and agent-authored
payloads with the locale-default codec:
- docx/powerpoint validators/base.py opened OOXML part XML in text mode
before handing it to lxml. On Windows (cp1251/GBK) the bytes decode to
mojibake that lxml then parses, so validation runs against silently
corrupted document text; on locales where the UTF-8 bytes don't decode
the validator crashes with UnicodeDecodeError instead of validating.
Opening as bytes lets lxml honor the encoding declared in the XML prolog.
- The pdf form scripts (fill_fillable_fields, fill_pdf_form_with_annotations,
create_validation_image, check_bounding_boxes) read the fields JSON the
agent authors — UTF-8 by construction — with the locale codec, so
non-ASCII form values (any Cyrillic/CJK/accented input) either crash or
get written into the user's PDF as mojibake. The json.dump writers use
ensure_ascii=True and were already safe; only the readers needed pinning.
Adds a contract test asserting every document/payload reader is
locale-independent, plus a live regression test that runs
check_bounding_boxes.py on a non-ASCII fields.json under a forced
non-UTF-8 locale — it fails without the fix on both POSIX (C locale)
and Windows (cp1251 chokes on the 0x98 byte of U+2018).
_setup_worktree read both files with the locale default encoding. On a
cp1251/GBK Windows machine a UTF-8 include list either decodes to
mojibake paths (non-ASCII entries silently not copied) or raises
UnicodeDecodeError, which the enclosing handler logs at DEBUG and
swallows — no include is copied at all, so the worktree starts without
.env/keys and the agent breaks invisibly. A Notepad BOM likewise glues
to the first include entry on every platform, and to the first
.gitignore line, defeating the '.worktrees/' membership check and
appending a duplicate entry on each run.
Read both files with utf-8-sig + errors=replace, matching the canonical
.env readers in hermes_cli/config.py (utf-8-sig because Notepad adds a
BOM) and the UTF-8 append this same block already performs on
.gitignore.
Regression tests exercise the real cli._setup_worktree: the two BOM
tests fail without the fix on any platform, the non-ASCII include test
additionally reproduces the Windows locale failure.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
_read_hub_installed_names() reads ~/.hermes/skills/.hub/lock.json with a
strict utf-8 decode. Hub skill descriptions can carry Windows-1252
typographic bytes (em-dash 0x97, smart quotes, bullets) as single high
bytes; read_text(encoding="utf-8") then raises UnicodeDecodeError, which
is a ValueError sibling not caught by the function's
except (OSError, json.JSONDecodeError). It escapes and 500s the whole
/api/skills endpoint, blanking the desktop Skills panel.
Decode with errors="replace" so the offending byte degrades to U+FFFD
and the structurally valid JSON — and every other skill — stays readable.
Fixes#68053
`_render_distribution_plan` reads the target profile's `.env` to decide
whether a required env var is already set (so it doesn't nag the user),
using `Path.read_text()` with no encoding. Two bugs:
1. `Path.read_text()` defaults to the system locale (cp1251/GBK on Windows),
which raises `UnicodeDecodeError` on any non-ASCII byte. The surrounding
`except OSError` does NOT catch that — `UnicodeDecodeError` is a
`ValueError` — so a mis-encoded `.env` aborts the entire install preview.
2. Even on a UTF-8 locale, a Notepad-added BOM prefixes the first key
(`KEY`), so the very first required env var is mis-reported as
"needs setting" when it is actually present.
`.env` is written as UTF-8 everywhere in the codebase. Read it as
`utf-8-sig` (tolerates the BOM) and also catch `UnicodeDecodeError` so a
genuinely un-decodable file skips the pre-check instead of crashing.
Regression tests: a BOM-prefixed `.env` whose first key must still read as
"set", and an invalid-UTF-8 `.env` that must not abort the preview.
Follow-up to review feedback:
- mem0 _prompt_api_key read .env with the locale default, so a Notepad
BOM hid the first key from the masked current-value lookup; read it
with utf-8-sig + errors=replace like the canonical readers in
hermes_cli/config.py.
- hindsight _load_simple_env used plain utf-8; it also parses the Hermes
.env during post_setup, where a BOM stuck to the first key. Switch to
utf-8-sig + errors=replace.
- Add hindsight regressions: BOM key matching in _load_simple_env and in
the cloud post_setup writer, plus non-ASCII round-trip preservation,
and a mem0 regression for the BOM'd masked-key lookup. The BOM tests
fail without the fix on any platform.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The mem0 and hindsight memory-provider setup routines round-trip the
user's ~/.hermes/.env: they read existing lines, update the keys they
manage, and rewrite the whole file preserving every other line verbatim.
Both used env_path.read_text() / write_text() with no encoding.
read_text()/write_text() with no encoding fall back to the system locale
(cp1252/GBK on Windows), so on a non-UTF-8 host the preserved lines get
mangled or the call crashes on any non-ASCII value, and — because the
reader never strips a BOM — a Notepad-edited .env makes the first key
fail the in-place match and get duplicated instead of updated.
Match the canonical .env readers in hermes_cli/config.py: read with
encoding='utf-8-sig' (BOM-tolerant) and write with encoding='utf-8'.
mem0/_setup.py already pins utf-8 for mem0.json, so this just aligns the
.env path in the same file. Fixes both memory plugins in one class fix.
Adds regression tests: a BOM'd .env updates the first key in place
(locale-independent, fails without the fix) and non-ASCII existing lines
survive the round-trip.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
On Windows with CJK locales (e.g. Chinese/GBK), pathlib.Path.read_text()
defaults to the system encoding instead of UTF-8, causing UnicodeDecodeError
when reading .env or .json config files that contain non-ASCII characters.
Explicitly pass encoding='utf-8' to all read_text() and write_text() calls
in the hindsight memory provider plugin.
Addresses hermes-sweeper review on PR #54866: the installer runs
tools/skills_sync.py as a child python.exe whose PYTHONIOENCODING /
PYTHONUTF8 the scoped install.ps1 block sets, but there was no
regression test for this child-Python UTF-8 path. The existing
test_child_process_inherits_utf8_mode covers a different (bootstrap
entry-point) flow.
Add TestSkillsSyncUtf8Guard: three subprocess tests that import
skills_sync (triggering its import-time stdout/stderr reconfigure)
and assert the checkmark/up-arrow glyphs the script prints at
tools/skills_sync.py:596,675 emit valid UTF-8 and exit 0 even when
the child env is left unset or explicitly hostile (gbk). A third
test proves the guard is load-bearing by reproducing the crash
without it.
Also keep the new install.ps1 comment ASCII-only (the checkmark
spelled out as U+2713) per the file's PS 5.1 parser-compatibility
contract at scripts/install.ps1:79-80; the literal glyph in the
comment violated that contract.
On Windows with a non-UTF-8 system locale (e.g. CP936/GBK on zh-CN),
Python defaults stdout/stderr to the active codepage. tools/skills_sync.py
prints glyphs such as checkmark (U+2713) and up-arrow (U+2191) that GBK
cannot encode, raising UnicodeEncodeError mid-run.
The installer (scripts/install.ps1) captures this script's stdout and the
Rust bootstrap parses it as UTF-8 expecting a JSON result frame. A GBK
byte stream (or the traceback it triggers) surfaces as:
WARN stdout read error: stream did not contain valid UTF-8
stage=config-templates state=Failed
error=install.ps1 -Stage config-templates produced no JSON result frame
(exit=Some(0))
i.e. the stage fails even though the script exits 0. install.ps1 already
sets [Console]::OutputEncoding = UTF8, but that does not propagate to the
python.exe child (Python reads PYTHONIOENCODING / locale, not the console
encoding).
Fix in two places for defense in depth:
- tools/skills_sync.py: reconfigure sys.stdout/stderr to UTF-8 at import so
output is valid UTF-8 regardless of caller or active codepage.
- scripts/install.ps1: set PYTHONIOENCODING=utf-8 and PYTHONUTF8=1 (scoped
to the call, restored afterwards) around the skills_sync.py invocation.
Path.read_text() without an explicit encoding uses the platform's
default encoding. On Windows this is typically cp1252 or mbcs, which
causes UnicodeDecodeError or silent data corruption when reading
UTF-8 content (JSON files, user text, config with non-ASCII chars).
This is the read-side companion to the write_text() encoding fix.
Fixed the most critical locations that read JSON data, user content,
and config files across 14 files with 31 call sites.
Pattern: .read_text() → .read_text(encoding='utf-8')
json.loads(path.read_text()) → json.loads(path.read_text(encoding='utf-8'))
Path.read_text() and Path.write_text() without explicit encoding
default to the system locale encoding. On Windows this is typically
cp1252, which causes UnicodeDecodeError for UTF-8 content (JSON
configs, user data, service scripts).
Add encoding="utf-8" to all read_text() and write_text() calls
across 8 CLI files, matching the pattern established in PR #50534
(security_audit_startup.py) and ruff rule PLW1514.
Fixed files:
- main.py: 4 read_text calls
- auth.py: 3 read_text calls
- banner.py: 1 read_text + 1 write_text
- service_manager.py: 1 read_text + 4 write_text
- container_boot.py: 1 read_text + 4 write_text
- doctor.py: 3 read_text calls
- uninstall.py: 2 read_text calls
- gateway.py: 1 write_text call
Several file-I/O call sites still use open() / Path.read_text() /
Path.write_text() without an explicit encoding, so they fall back to
the platform default. On Windows CN/JP/KR locales (GBK/CP932/CP949)
any non-ASCII byte in a config/state/user-content file raises
UnicodeDecodeError or UnicodeEncodeError and crashes the caller.
to the remaining hot paths:
- agent/copilot_acp_client.py: fs/read_text_file and fs/write_text_file
(Copilot's read_file / write_file tools,
directly reported in #18637 bug 2)
- agent/model_metadata.py: context-length YAML cache load + two
save sites (context probing is on the
call path of every model invocation)
- agent/nous_rate_guard.py: cross-session rate-limit JSON state
(read + atomic write via os.fdopen)
- cron/scheduler.py: user config.yaml read in run_job
- gateway/delivery.py: cron output writes for AI-generated
content, very likely non-ASCII
yaml.dump call sites also gain allow_unicode=True so the emitted
YAML preserves non-ASCII chars as-is instead of emitting \u escape
sequences.
Adds regression tests that monkeypatch builtins.open / Path.read_text
/ Path.write_text to simulate a GBK locale: each test raises
UnicodeDecodeError / UnicodeEncodeError unless the caller explicitly
passes encoding='utf-8'. Verified that the tests fail on main and
pass with this change, on Linux as well as on Windows.
Refs #18637
The three refresh tests asserted the replaced shared client gets
close()d — the exact cross-thread close#70773 removes. They now pin
the new contract: close() is NOT called from the refresh path; the
old client is retired (sockets shutdown, FD release deferred to GC).
Widen the #70773 fix beyond the three in-request cleanup sites removed in
the cherry-picked commit: every remaining path that swaps out the shared
OpenAI client could still hard-close its pool from a thread that doesn't
own the in-flight sockets (credential rotation/refresh on the turn thread,
dead-connection cleanup, gateway cache eviction, transport recovery) —
the same FD-recycle corruption vector, just rarer.
Add AIAgent._retire_shared_openai_client(): shutdown(SHUT_RDWR) all pooled
sockets (FD-safe from any thread, unblocks in-flight readers) but never
call client.close() — FD release is deferred to GC, which cannot run until
every borrowing thread has unwound its SSL BIO. Refcounting is the
ownership handshake; with no borrowers the FDs are released immediately.
Wired into:
- _replace_primary_openai_client (rotation/refresh/dead-conn cleanup)
- try_recover_primary_transport (primary_recovery)
- release_clients (gateway cache_evict)
agent.close() keeps the hard close: full teardown is a real session
boundary where no request may be in flight.
Tests: new tests/run_agent/test_70773_shared_client_fd_corruption.py
covers the three watchdog/retry sites plus retire semantics; existing
close-assertions updated to pin retire-not-close.
The streaming stale watchdog was calling
_replace_primary_openai_client() from its polling thread, which closes
the shared client's connection pool. Worker threads from previous
stale-killed attempts may still be unwinding their SSL BIOs, causing
TLS application-data to overwrite SQLite file headers via FD reuse.
This is the same corruption vector documented in #67142 for Anthropic,
where the fix was to never close the shared client from a non-owner
thread. Apply the same pattern to the OpenAI-wire path:
- Stale stream watchdog: skip shared client replacement
- Mid-tool-retry cleanup: skip shared client replacement
- Stream retry cleanup: skip shared client replacement
The request-local client is already closed via _close_request_client_once.
The shared client is replaced lazily by _ensure_primary_openai_client
on the next request, which runs on the owning thread.
Closes#70773.
CI's check:lint failed on two perfectionist rule violations introduced by the
new test file: type import ordering and missing blank line between the
parent-relative and same-directory import groups. No behavior change.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Two React passive-effect timing bugs let a session switch land in the wrong
chat: activeSessionIdRef/selectedStoredSessionIdRef (use-session-state-cache)
and the composer's attachment-scope swap (use-composer-draft) both mirrored
their source props via useEffect, which fires one commit AFTER the new
session's view has already painted — a synchronous read/submit in that window
observed the outgoing session's ids/attachments.
- use-session-state-cache.ts: mirror the session refs synchronously during
render instead of a useEffect, guarded to fire only when the prop itself
changed (not unconditionally) so an imperative pin from submit.ts /
use-session-actions (e.g. a freshly resumed runtime id, intentionally not
synced to the source atom) survives an unrelated re-render.
- use-composer-draft.ts: the per-thread attachment-scope-swap effect is now a
useLayoutEffect, closing the window before paint.
- submit.ts / session-context-drift.ts: add a 3rd drift prong comparing the
composer's loaded scope (SubmitTextOptions.composerScope) against the
submit target, resolved into the same lineage-root domain
(resolveComposerSessionKey) the composer itself uses — comparing against
the raw tip id would false-positive-abort every submit into any session
that has ever auto-compressed.
- routes.ts / chat/index.tsx: the primary composer's durable scope key now
prefers the route over a possibly-stale store selection
(primaryRouteSelectedSessionId).
- use-composer-draft.ts: redacted [composer-rehydrate] diagnostic log
(counts/kinds/scope only, never raw refs) for future reports in this class.
- chat-runtime.ts: normalize attachment id values (url/path) before hashing
so a re-attach with a trailing slash or backslash path dedupes correctly.
16 files, 286 tests across the touched/dependent suites (17 files) green,
including new regression coverage for each fix.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Teardown-path tests build bare runners via object.__new__ without
the liveness-guard machinery; the unguarded call raised
AttributeError in 8 tests. Same guard pattern as the start path.
Follow-up to the salvaged #69164 commits: policy forbids introducing new
HERMES_* environment variables, so the four watchdog env knobs
(HERMES_GATEWAY_LOOP_WATCHDOG / _INTERVAL / _TIMEOUT / _STRIKES) are
replaced with a single config.yaml boolean:
gateway:
loop_watchdog: true # default; false disables both guards
- gateway/config.py: new GatewayConfig.loop_watchdog field (default True),
parsed from top-level or nested gateway: form, round-trips via
to_dict/from_dict.
- gateway/run.py: _start_loop_liveness_guards() checks config.loop_watchdog
before arming the floor timer + watchdog (getattr-guarded for bare
object.__new__ runners).
- gateway/shutdown_watchdog.py: start_loop_liveness_watchdog() no longer
reads the environment; probe interval/timeout/strikes are module
constants (30s/10s/3 — ~90s to restart, matching the systemd watchdog
layer's posture).
- hermes_cli/config.py: documented gateway.loop_watchdog default so
'hermes config set gateway.loop_watchdog false' validates.
- tests: env-knob tests replaced with config-gate + round-trip tests;
the final-strike boundary test injects its probe via max_strikes
directly instead of patching the removed env helper.
- A stop() landing while the final diagnostics (critical log,
traceback dump) are executing could still reach os._exit(75) after
the pre-diagnostic check. Add a third stop_event recheck immediately
before the hard exit: diagnostics may complete, but a disarmed
watchdog never exits.
- Deterministic regressions for both windows (stop triggered from
inside logger.critical and from inside faulthandler.dump_traceback);
mutation-verified (removing the check turns both red). Frozen-loop
semantics unchanged.
Addresses the second round of the shutdown-race review on #69164.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- Re-check stop_event after a missed probe (before the strike
increment) and again on entering the final-strike branch (before the
critical log, dump, and hard exit), so a normal stop() landing
between the last timeout check and the exit path can no longer be
misclassified as a freeze and trigger a supervisor restart.
- Deterministic boundary tests pin both re-checks independently
(mutation-verified: removing either check turns its own test red);
frozen-loop semantics are unchanged.
Addresses the shutdown-race review on #69164.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- A self-rescheduling 5s call_later floor timer, armed before any
adapter connects, guarantees the selector always has a finite
timeout, so the existing async defenses (polling heartbeat, timeout
guards) regain a chance to run after a zero-pending-timer stall.
- A resident daemon-thread liveness watchdog probes the loop via
call_soon_threadsafe every 30s; after 3 consecutive 10s-timeout
misses (~120s of total unresponsiveness) it dumps all thread
tracebacks and exits with the established
GATEWAY_SERVICE_RESTART_EXIT_CODE (75) so a supervisor restarts the
gateway - async-level recovery cannot run on a frozen loop.
- stop() disarms both guards before any teardown await so a busy
shutdown is never misjudged as a freeze.
HERMES_GATEWAY_LOOP_WATCHDOG=0 disables; _INTERVAL/_TIMEOUT/_STRIKES
tune the thresholds.
Fixes#69089
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
signal.SIGUSR2 and faulthandler.register() don't exist on Windows;
the bare reference raised AttributeError at import time per the
windows-footgun checker. faulthandler.enable() still covers
fatal-error dumps on all platforms.
When connected_count == 0 and at least one platform failed with a
non-retryable error, the runner exited with GATEWAY_FATAL_CONFIG_EXIT_CODE
(78) even if OTHER platforms failed for merely transient reasons.
Real-world shape (NS-609, hosted instance): WhatsApp enabled but never
paired (non-retryable whatsapp_not_paired) + Telegram TimedOut during
polling startup (retryable) => exit 78 => the gateway either goes
permanently down (supervisors honoring the exit-78 contract via
RestartPreventExitStatus / the s6 finish->125 translation from #51228) or
crash-loops (anything else). Either way Telegram never gets its retry and
the dashboard drops with every exit, so a single unpaired platform plus
one network blip disconnected every channel on the instance.
Now exit 78 is reserved for the case where ALL startup failures are
non-retryable (true config error, nothing to wait for). With mixed
failures the gateway stays alive in degraded state: the reconnect watcher
recovers the retryable platforms and the misconfigured ones stay
fatal-parked and visible in runtime status.
Three-part fix for the gateway going silently deaf after a retryable
fatal adapter error (e.g. httpx.ConnectError on Telegram):
1. **Detach-on-timeout in _connect_adapter_with_timeout** — Replaced
plain asyncio.wait_for with the task-detach pattern used by
_await_adapter_cleanup_with_timeout. asyncio.wait_for cancels the
overdue task but then waits for it to exit, so a connect() that
catches CancelledError can block recovery forever. The detach
pattern releases the runner at the deadline via
consume_detached_task_result.
2. **Ensure reconnect watcher always runs after escalation** — Added
_ensure_reconnect_watcher_running(), called after queueing a
retryable fatal error. If the reconnect watcher task has died
(exhausted restart budget, terminal exception), it is respawned
so queued platforms are never permanently stranded.
3. **Faulthandler at gateway startup** — Enabled faulthandler +
SIGUSR2 dump to a rotating file under HERMES_HOME/logs/ for
post-mortem diagnosis of future event-loop freezes.
Tests added for _ensure_reconnect_watcher_running (alive, dead,
not-started, not-running), fatal-error integration (retryable calls
ensure, non-retryable does not), and _connect_adapter_with_timeout
(timeout raises, success returns).
/api/status?profile= now passes pid_path=/path=/expected_home= to the
PID and runtime-status readers; the profile-unification fakes had
zero-arg signatures and raised TypeError. Plain /api/status call shapes
are unchanged (pinned by the existing zero-arg tests in
test_web_server.py).
Follow-up for the salvaged #70498 fix: replace the original PR's
mock-signature churn (28 lambda **kw edits, needed only because it changed
the no-profile call shape) with two targeted regression tests:
- ?profile=<name> must pass the profile's gateway.pid / gateway_state.json
paths and expected_home to the gateway status readers (HOME-anchored
per-profile state under ~/.hermes/profiles/<name>/)
- ?profile=<unknown> must 404 via _resolve_profile_dir
The production change keeps plain /api/status on the exact zero-arg calls,
so every pre-existing test passes unmodified.
When ?profile=<name> was passed to /api/status, the handler used
_config_profile_scope to set the HERMES_HOME contextvar override, but the
gateway liveness check (get_running_pid_cached) and runtime status read
(read_runtime_status) both resolve _get_process_hermes_home(), which
deliberately ignores contextvar overrides (issue #56986) — it always reads
os.environ['HERMES_HOME'] or the platform default. A named profile's
gateway identity files (~/.hermes/profiles/<name>/gateway.pid,
gateway_state.json) were therefore never found and the endpoint always
reported the profile's gateway as stopped.
Fix: when ?profile=<name> is requested, resolve the profile directory and
pass explicit profile-scoped paths:
- get_running_pid_cached(pid_path=profile_dir / 'gateway.pid')
- read_runtime_status(path=profile_dir / 'gateway_state.json')
- get_runtime_status_running_pid(..., expected_home=profile_dir)
This is the same explicit-path pattern _collect_profile_gateway_topology
already uses for per-profile gateway state, and it works within the #56986
constraint (no HERMES_HOME env mutation; read-only cross-profile access).
Plain /api/status without ?profile= keeps the exact zero-arg calls, so its
behavior — including the pid-cache signature and runtime-status fallback —
is byte-for-byte unchanged.
Fixes#69143
The .env sanitizer inferred missing newlines from known KEY= substrings
inside existing values. Plain secrets containing those bytes could therefore
be split into synthetic assignments and rewritten to disk.
Treat each physical line as the only assignment boundary and keep bytes after
the first equals sign opaque for boundary discovery. Preserve safe formatting,
null-byte removal, BOM handling, and normal one-assignment-per-line parsing.
Cover direct loading, dotenv loading, sanitization, writers, and migration
with behavioral regressions.
Fixes#29155
Follow-up to the cherry-picked #68258 base: the cross-session-stable
prefix (_cached_system_prompt_static) was only recorded on fresh
builds, so two paths silently degraded to the legacy single-breakpoint
layout (flagged in review of #68258/#69341/#69704):
- Session restore: gateway surfaces build a fresh AIAgent per turn and
restore the persisted prompt verbatim from the session DB; the static
prefix stayed None from turn 2 onward, flip-flopping the wire layout.
- Post-compression cached-prompt reuse: _invalidate_system_prompt()
clears the static prefix, and the keep-cached-prompt branch never
restored it.
Both sites now reconstruct the stable tier and adopt it ONLY when the
authoritative prompt string literally startswith() it — stable-tier
drift (skills edited, identity changed) falls back to the legacy layout
with the stored bytes untouched. Fail-open on any builder error. The
restore-path rebuild is gated on _use_prompt_caching so non-Anthropic
routes skip it entirely.
Refs #68191
Co-authored-by: JonthanaHanh <92574114+JonthanaHanh@users.noreply.github.com>
Co-authored-by: joaomarcos <joaomarcosdias444@gmail.com>
Co-authored-by: webtecnica <webtecnica@users.noreply.github.com>
The conversation loop normalizes message text right before the API call so
the request prefix is byte-identical across turns -- the stated reason is
KV cache reuse on local inference servers and better cache hit rates on
cloud providers. Cache breakpoints were injected *before* that pass, which
defeats it.
`_apply_cache_marker` rewrites a plain-string `content` into a
`[{"type": "text", ...}]` block. The normalization pass is guarded on
`isinstance(content, str)`, so every message that just got marked is
silently skipped by it and keeps its raw leading/trailing whitespace. A
message is only marked while it sits in the last-3 window, so:
turn N in the window -> marked, content "file1\nfile2\n"
turn N+1 rolled out -> plain, content "file1\nfile2"
The same logical message is sent with different bytes on consecutive
turns. The prefix stops matching at that position -- which is inside the
span the breakpoints were placed to protect -- so the reusable prefix
collapses back toward the system breakpoint on every turn. Tool results
carry a trailing newline almost by default (any shell command output), so
this is the common case, not an edge case.
Move the injection below every message mutation. Besides fixing the
whitespace divergence this stops breakpoints from being spent on messages
that the orphan sweep or the thinking-only drop is about to remove or
merge away -- a marker on a dropped message is a wasted breakpoint out of
the four available.
Nothing between the old and new call sites reads `cache_control`, and the
mutators now see the plain-string shapes they were written against.
Follow-up for salvaged PR #69141, addressing the last open review point:
cmd_prune() only set orphan_allowlist inside 'if orphans or pre_v2_orphans',
so a zero-orphan preview passed the unrestricted None sentinel down to
prune_checkpoints(), authorizing deletion of any project that became
orphaned between the preview and the rescan — with zero confirmation
calls. The allowlist is now bound unconditionally for every non-force
run (empty preview => empty allowlist); --force keeps None. Adds the
zero-orphan-preview timing regression plus allowlist-identity tests.
Address P1 from PR review: cmd_prune()'s y/N preview reads
store_status() but the confirmed deletion re-scans both the v2 and
pre-v2 layouts from scratch. A workdir that goes missing while the
human is answering the prompt gets swept in as if it had been shown
and approved.
prune_checkpoints() now accepts orphan_allowlist — a set of v2 project
hashes and/or pre-v2 shadow repo paths. When set, only orphans whose
identity is in the set are deleted; anything newly orphaned since the
scan survives the run. cmd_prune() builds this set from the exact
projects it just displayed and passed confirmation for. --force still
passes None (no preview shown, so nothing to bind to).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Requested by egilewski on #69141: the orphan confirmation flow had no
test coverage at all before this. Exercises hermes_cli.checkpoints.cmd_prune
directly against pre-v2-only and mixed (v2 + pre-v2) fake stores —
decline aborts with nothing deleted, accept deletes both layouts,
--force and --keep-orphans skip the prompt as expected.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
store_status()["projects"] only ever covered v2 metadata, so the
`hermes checkpoints prune` confirmation prompt was blind to pre-v2
base/<hash>/HEAD shadow repos that prune_checkpoints() deletes
separately via shutil.rmtree — a pre-v2-only or mixed store could
lose checkpoint history without ever hitting the confirmation.
Extract the pre-v2 scan into _pre_v2_shadow_repos() and have both
store_status() (preview, new pre_v2_projects key) and
prune_checkpoints() (deletion) read from it, so the CLI prompt can
no longer diverge from what actually gets removed.
Addresses review from egilewski on #69141.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Builds on this PR's diagnosis by @Frowtek: a missing workdir is
ambiguous (deleted project vs. an unmounted external volume / network
share / VPN not yet up), so it's not safe evidence for a destructive
GC sweep — especially one that runs unattended at startup.
- cli.py / gateway/run.py: the startup auto-maintenance sweep now
always passes delete_orphans=False to maybe_auto_prune_checkpoints().
It still prunes by retention_days, size cap, and legacy archives —
none of which require guessing whether a project was deleted or is
just temporarily unreachable.
- hermes_cli/config.py: drop the now-unused delete_orphans default.
- hermes_cli/checkpoints.py: `hermes checkpoints prune` (the explicit,
human-invoked path) now previews the orphan project list and asks
for confirmation before deleting, unless -f/--force is passed.
- Docs updated (EN + zh-Hans) to match.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Two compression-tip hydration tests simulated legacy state by emptying
the parent AFTER end_session(compression) — exactly the durable write
the new closed-parent guard refuses. Reordered: empty first, close
second. The tests' actual contract (old id hydrates from the live tip)
is unchanged and still pinned.
The salvaged drift check compared durable rows to the in-memory snapshot
by content and ran in both modes. Two problems:
1. In-place compaction (the default) archives non-destructively — drift
cannot lose data there, and the strict-prefix content comparison
failed against seeded histories, aborting every in-place compaction
(5 test failures in test_in_place_compaction.py).
2. Content equality wedges on sessions with legal in-memory mutation of
past turns (multimodal compression, retry replacement) — the same
permanent-abort shape as #14694.
Now rotation-only and length-based: abort only when the durable parent
has MORE rows than the snapshot (a writer committed in the lease window).
Dead helper _durable_history_matches_snapshot removed.
The runtime-repair change passes repair_observer= to update_managed_uv/
ensure_uv; the autouse fixture fakes had zero-arg signatures and raised
TypeError through the mock. Sibling test file to the PR's own suite.
Follow-up on the #70186 salvage. The cherry-picked repair pinned the
candidate to the exact current CPython patch (e.g. 3.11.14). Verified
live with uv 0.11.19: every published python-build-standalone artifact
for 3.11.14 links vulnerable SQLite 3.50.4 — even with --reinstall — so
the exact-patch pin made the repair permanently impossible on the
installs that need it most (repair_vulnerable_runtime returned
'failed: could not provision a fixed private Python runtime').
Request the minor line (3.11) instead — the same resolution a fresh
'uv python install' would make, still inside requires-python — and
tighten the drift gate to 'same minor, no downgrade'. E2E-verified
end-to-end on a real vulnerable venv: repair_vulnerable_runtime()
provisioned 3.11.15, built + smoke-tested the sibling venv, cut over,
and reported SQLite 3.50.4 → 3.53.1 with the old venv parked for
rollback.
Follow-up fixes on top of the cherry-picked guard:
- verify_sqlite_integrity(): an oversized (max_bytes-exceeding) database
now still fails on a zeroed/invalid header — previously valid=True was
set before the header check, so a >1GiB zeroed state.db (the exact
#68474 signature at 95MB scale) passed as valid.
- _run_pre_update_backup(): the guard referenced get_hermes_home before a
later function-local 'from hermes_constants import get_hermes_home'
shadowed it → UnboundLocalError swallowed by the snapshot try/except,
silently disabling the post-snapshot check AND the snapshot-id output.
Alias the import explicitly.
- Import _quick_snapshot_root where used (was NameError in all 3 guards).
- tests/hermes_cli/test_state_db_guard.py: real-SQLite E2E coverage —
valid/zeroed/truncated files, oversized header gate, copy+verify
roundtrip, snapshot restore flow, and the live _run_pre_update_backup
path against a temp HERMES_HOME with mid-flight zeroing.
Problem:
On Windows, state.db could be silently replaced with 95MB of null bytes
during a desktop update (v0.19.0). The pre-update snapshot was valid, but
the live file was destroyed and the update reported exit code 0, masking
the data loss. Sessions between the snapshot and the update were
irrecoverable.
Root cause analysis:
The update flow (Desktop Electron → hermes-setup.exe → hermes update)
kills the backend process tree via taskkill /T /F, then pauses Windows
gateways, creates a pre-update snapshot, runs git pull + pip install, and
resumes gateways. On Windows, a force-killed process holding state.db
(SQLite WAL mode) can leave the file open to races with antivirus/NTFS
filter drivers, or the gateway resume can encounter a partially-recovered
WAL state that results in a zeroed file — all while exit code 0 reports
success.
Fix — three layers of defense:
1. Emergency desktop-side backup (pre-flight):
- New function in Electron main.ts reads the
SQLite header, logs it, and takes a timestamped emergency copy of
state.db BEFORE the backend is killed or the updater is spawned.
Runs in both the Tauri-updater path (Windows) and the in-app update
path (Posix). Prunes to the 2 most recent emergency backups.
2. Pre-update integrity verification (Python CLI):
- After creates the pre-update snapshot,
checks the LIVE state.db file (header +
PRAGMA integrity_check). If corrupted, checks whether the snapshot
copy is valid and warns the user. The update still proceeds because
the snapshot is the recovery path.
3. Post-update auto-restore (Python CLI):
- After the update completes (both git-pull and ZIP paths), verify
state.db integrity. If corrupted/zeroed, automatically restore from
the pre-update snapshot and re-verify. This catches the exact case
where state.db was destroyed mid-update but the snapshot was valid.
New functions in hermes_cli/backup.py:
- verify_sqlite_integrity(path, check_header, run_pragma, max_bytes)
→ Three-stage check: file size, SQLite header magic, PRAGMA
integrity_check. Configurable max_bytes to avoid reading huge DBs.
- copy_db_and_verify(src, dst)
→ Like _safe_copy_db() but verifies the destination after backup.
Fixes#68474
When a background terminal() command backgrounds its own long-lived
child (`node server.js &`, `sleep 300 &`), the grandchild inherits the
write end of the reader thread's stdout pipe. The direct bash child
exits promptly, but the pipe never reaches EOF while the grandchild
lives — so `_reader_loop`'s blocking `read1()` parked the thread
forever, `session.exited` never flipped on its own, and
`notify_on_complete` was silently lost. `_reconcile_local_exit`
(#17327) only runs lazily from poll()/wait(), so nothing autonomous
ever surfaced the exit; each occurrence also leaked a reader thread
and pipe fd for the grandchild's lifetime.
Fix: on POSIX, drain via select() with a short poll interval and stop
shortly after the direct child exits even if the pipe hasn't EOF'd —
the same pattern the foreground path uses in
tools/environments/base.py::_wait_for_process (#8340). Windows pipes
don't support select(), so the blocking path is kept there with the
existing lazy reconcile as the safety net; mocked/iterator stdout
streams (no usable fileno) also keep the historical path.
Fixes#68915
With the eager pre-warm removed (PR #66783), slash.exec is the only spawn
path — and it runs on the RPC thread pool, so two concurrent worker-routed
commands on a fresh session could both see slash_worker=None and each fork
a full stdio-MCP-fleet worker (the _attach_worker race loser leaking
unclosed). Add a per-session spawn lock with a double-check, plus a
regression test racing two slash.exec calls through handle_request.
Also maps Ne0teric's contributor email.
Every slash_worker child runs its own MCP discovery (#61891), which
forks the full configured stdio MCP fleet — on a config with a handful
of stdio servers that is ~20 OS processes per worker once npx/cmd
wrappers are counted. The gateway pre-warmed a worker for every session
at create/build time, and sessions held by a live transport are (by
design) never reaped, so a desktop app left open for days accumulates
one fleet per retained session. On a real setup this reached ~120
processes across 6 sessions and pushed Windows commit charge to the
point where CreateProcess started failing system-wide ("Not enough
memory resources are available to process this command").
slash.exec already spawns a worker on demand when the session has none
and already recovers from a dead worker the same way, so the eager
pre-warm is pure pre-warming:
- drop the pre-warm in the deferred session-build path
- drop the pre-warm in _init_session
- make _restart_slash_worker a no-op for sessions that never spawned a
worker (the next slash.exec builds one with the current session
key/model, so no stale-key worker can exist)
Only sessions that actually run a worker-routed slash command now pay
for a fleet. Cost: the first such command in a session takes the CLI
build + MCP discovery hit that session.create used to absorb.
Tests: the two create/close-race guards now assert the build thread
never constructs a worker (the notify-unregister guarantees are kept);
the restart-orphan guard seeds a live worker so the close path is still
exercised; new test pins the restart no-op for workerless sessions.
Follow-up hardening on top of the salvaged #69745 guard, addressing both
review findings:
- Drift detection no longer re-reads the file. _reload_target performs ONE
checked read and derives both the drift check and the entry parse from that
same raw snapshot (_detect_external_drift now takes the raw text). The old
second read swallowed OSError as 'no drift', so a read failure between the
two reads let replace/remove/apply_batch rewrite the file from a stale view,
discarding externally added entries.
- Invalid UTF-8 now counts as unreadable: the checked read catches
UnicodeDecodeError and mutations return the preservation refusal instead of
raising (or worse, rewriting bytes we can't round-trip).
- USER.md is covered by the same guard (shared _reload_target path) and now
pinned by an explicit test.
Tests: read-once structural invariant, invalid-UTF-8 refusal with
byte-identical file, user-store refusal.
`_read_file` degraded any read failure to `[]`, conflating "file exists but
couldn't be read" with "empty store". That is a silent, total data-loss bug on
the `add` path.
`add` re-reads the file under lock, appends the new entry, and rewrites the
WHOLE file from the parsed entries. It deliberately skips the drift guard
(#42874: "appending never clobbers existing content") — but that reasoning only
holds when the reload actually saw the file. When `read_text` raises
(an external editor momentarily holding the file on Windows, a permission
change, a filesystem/EINTR blip), `_read_file` returns `[]`, so `add` treats
the store as empty and rewrites the file down to just the new entry — every
prior memory gone — while returning `success: True`.
Reproduced with a transient read failure during `add`:
entries on disk before : 3 (dark-mode pref, deadline, deploy target)
add("A brand new fact") : success=True
entries on disk after : 1 ("A brand new fact") <-- the other 2 wiped
replace/remove/apply_batch were shielded only incidentally — an empty view
means `old_text` never matches, so they abort before writing — but they still
returned a misleading "no entry matched" instead of naming the real problem.
Fix: distinguish unreadable from empty. `_read_entries_checked` returns
`(entries, read_ok)`, with `read_ok=False` only when the file exists but can't
be read; absent/empty stays a clean `([], True)`. `_reload_target` returns a
`_READ_FAILED` sentinel in that case without touching in-memory state, and all
four mutation paths (add, replace, remove, apply_batch) refuse the write with a
clear "retry in a moment" error. This is the same posture as the drift guard
and the pairing/checkpoint fixes: never rewrite a file from a view that isn't
the real one. `_read_file` keeps its `[]`-on-error contract for the read-only
`load_from_disk` caller, which never persists.
tests/tools/test_memory_tool.py: new TestUnreadableFileDoesNotWipeMemory —
add/replace/remove/apply_batch all refuse and leave the file byte-identical on
a transient read failure, plus controls that an absent file is still a clean
empty store and the happy path is undisturbed. The four refusal tests fail on
main. Full suite: 90 passed, 1 pre-existing failure (`test_deduplication_on_load`,
a UnicodeDecodeError unrelated to this change, identical on clean main).
Named endpoints from the providers: mapping (and legacy custom_providers:
list) never appear in the ACP model selector: _build_model_state lists
only the canonical current provider's catalog, and canonical provider
enumeration does not include user-defined named endpoints. The TUI
/model picker already renders these entries (#47039, implemented for the
TUI surface only), so editor clients silently hide endpoints the user
configured — e.g. an OpenAI-compatible Bedrock Mantle Responses provider.
Add _named_custom_provider_catalogs(), sourcing entries from
get_compatible_custom_providers() (covers both config shapes), and append
its models to the selector payload. Choice ids use the custom:<name>
slug shape so custom:<name>:<model> selections round-trip through
parse_model_input / resolve_runtime_provider unchanged on set_session_model.
Declared models (default_model + models) survive failed live /models
discovery — some OpenAI-compatible endpoints expose no /models route yet
serve their declared models fine. Honors providers.<name>.enabled: false
and discover_models: false.
Verified: scripts/run_tests.sh tests/acp/ — 318 passed, 0 failed;
scripts/check-windows-footguns.py clean.
When two consecutive compactions each failed to clear the threshold, the
anti-thrashing breaker blocked automatic compaction PERMANENTLY for the
life of the session: nothing decremented _ineffective_compression_count
(or _fallback_compression_streak) while blocked, so a session whose
middle region was briefly too small to compact never auto-compacted
again — it grew unbounded until the provider's hard context limit, and
only /new or /reset recovered it.
Recovery is a probation probe, not amnesty: after
_ANTI_THRASH_RECOVERY_SECONDS (300s) of continuous block the gate grants
exactly ONE attempt by dropping tripped counters to 1 strike (persisted,
so sibling agents on the same session row — gateway hygiene — unblock
too). An ineffective probe re-trips the guard on the next real-usage
verdict and the next recovery waits a full fresh window, so the worst
case in a truly incompressible session is one compaction attempt per
window — bounded, not thrash.
The recovery clock is armed lazily on the first BLOCKED evaluation and
is deliberately not durable: a restart that loads a durable tripped
counter (#69872) starts a full fresh window blocked, preserving the
restart-must-never-disarm contract (#54923).
Fixes#14694
Review follow-up for the dropped tool-call recovery (#69630): the
re-prompt pair was tagged _dropped_toolcall_nudge, but that marker was
not part of the ephemeral-scaffolding contract. _persist_session /
_flush_messages_to_session_db would therefore write the synthetic
'issue the actual tool call now' user message (and the narration-only
interim assistant turn) as real transcript rows — a resumed session
could replay the internal retry instruction as user-authored context
and prompt unsolicited tool use.
- Add _dropped_toolcall_nudge to _EPHEMERAL_SCAFFOLDING_FLAGS
(run_agent.py) so both SQLite and JSON persistence skip the pair.
- Add it to _SYNTHETIC_USER_FLAGS (conversation_compression.py) so the
compressor never treats the nudge as human intent.
- Flag the interim assistant half of the pair too, and include the
marker in the finalization scaffolding pop so a genuine turn end
strips the pair from the live transcript (mirrors the
_empty_recovery_synthetic pattern).
- Regression tests: flagged messages classify as ephemeral, the
returned transcript contains no scaffolding, and the turn tail stays
on the real assistant answer.
The initial fix guarded the no-tool-calls else branch, but that branch only
SETS final_response — the turn actually finalizes later, in a separate block
after final_msg is built. Runs that reached finalization via that path exited
without the guard ever running (observed live: a scheduled PR reviewer stalled
at tool_turns=1-2 with zero recovery nudges).
Move the recovery to the finalization chokepoint (right after final_msg is
built), so it catches every path that ends a turn. Single guard now:
- increments a consecutive-stall counter and re-prompts (bounded to 3),
- resets on any successful tool round, and
- resets on a genuine (non-mismatch) turn end,
so it guards each stall independently without capping the whole run and
without looping forever.
Verified live: the scheduled reviewer now recovers through the stalls and
submits real reviews — PR 57800 APPROVED, PR 54826 COMMENTED — one PR per run.
Some providers (observed: claude-opus-4.8 / claude-sonnet-4.5 on GitHub
Copilot, ~2026-07) return finish_reason="tool_calls" while the parsed
tool_calls array is empty — the model signalled it wanted to act but the
payload shipped no call. The conversation loop took the no-tool-calls
else branch, treated the turn's narration as the final answer, and exited
with the task unstarted.
On unattended multi-step jobs this is silent failure: a scheduled PR
reviewer, for example, would narrate "Let me verify the PR..." and stop
at tool_turns=0 every run, never submitting a review, while the job still
reported success.
Fix: in the no-tool-calls branch, detect the provider contract violation
(finish_reason == "tool_calls" with zero tool_calls) and re-prompt the
model to emit the call instead of exiting. Bounded to 3 consecutive
stalls; the budget resets after any successful tool round so it guards
each stall independently rather than capping the whole run. The narration
may live in content or only in the reasoning field (empty content) — the
guard keys on the finish_reason/tool_calls mismatch, so both are covered.
A genuine finish_reason="stop" text turn is unaffected.
Verified live: a scheduled reviewer that died at tool_turns=0 every run
now recovers through 20+ stalls per run and submits real reviews.
Tests: tests/run_agent/test_dropped_tool_call_recovery.py covers the
re-prompt, the empty-content case, the clean-stop control, and the
bounded-loop guard.
The last_activity field showed 'tool completed: read_file (120.7s)'
identically whether the tool succeeded or failed, making post-mortem
analysis of silent-hang reports (#69131) unnecessarily hard.
Append ' (error)' to the activity description when the tool result is
classified as a failure, in both the concurrent and sequential
execution paths.
(Salvaged from PR #69467; its conversation_loop.py activity-touch hunk
is the same fix as PR #69577 and lands in the preceding commit.)
After a tool call completes, the conversation loop does to
start the next API call. Between the last (from tool
completion in tool_executor.py) and the next one (at the start of the
next API call), there is a gap. If anything during this gap takes time
— context compression, slow provider prefill, or other post-tool
processing — the combined inactivity window can exceed the gateway's
inactivity_timeout (default 120s). The gateway kills the session and the
user sees 'agent never returns a final response', even though the tool
call itself succeeded in 0.1-0.2s.
Fix: call right before so the gateway
sees a fresh timestamp immediately after tool results are posted,
regardless of how long the follow-up API call takes.
Fixes#69559
Eliminates the separate import/set/clear dance for _SESSION_CWD by
passing cwd= directly to set_session_vars(), which already handles
the ContextVar set internally and clears it via clear_session_vars().
Also includes the exception message in the no_agent error path.
Follow-up to salvaged PR #70548 (#69396).
The cron scheduler was mutating process-global state in two places:
1. no_agent path called os.chdir() which changed the global process cwd,
leaking into concurrent gateway sessions.
2. The agent path set os.environ['TERMINAL_CWD'] which any gateway
session could read during context-file discovery via
resolve_context_cwd/build_context_files_prompt.
Fix:
- no_agent path: pass workdir as subprocess cwd parameter to
_run_job_script() instead of os.chdir(). The Python process cwd
is never mutated.
- Agent path: in addition to the lock-serialized TERMINAL_CWD,
also set the per-context _SESSION_CWD ContextVar from
agent.runtime_cwd. This ContextVar is scoped to the current
thread/context and NEVER leaks into other sessions.
resolve_context_cwd() checks _SESSION_CWD first, so the cron's
own context file discovery uses the correct workdir, while
gateway sessions (which have no override) fall through to their
own TERMINAL_CWD.
Sibling of the #69678/#69567 ledger leak class found while widening the
sweep: _probe_state_db used 'with sqlite3.connect(...)', whose context
manager only commits/rolls back and never closes, leaking one connection
(db fd) per health poll in the long-running gateway. Wrap the connection
in contextlib.closing so every probe closes deterministically.
Three durable ledgers used `with _connect() as conn:` where the sqlite3
connection context manager commits/rolls back but never closes, leaking the
db/-wal/-shm file descriptors on every call. On a long-running gateway this
exhausts RLIMIT_NOFILE and fails unrelated components with
`[Errno 24] Too many open files`. Same bug class as the cron execution ledger
(#69567 / PR #69594), which the connection helpers here are modeled on.
Fix: route every ledger operation through a `_transaction()` context manager
that guarantees `conn.close()` on exit. `_connect()` keeps its
schema-on-connect contract (several tests call it directly) and now self-closes
if schema init fails.
Adds per-module regression tests asserting every opened connection is closed,
including the no-op-update and exception-mid-transaction paths.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Follow-up to the cherry-picked #68633 commits, closing the final open
review point (egilewski): _relocated_replay_cache_control was applied
only inside `if replayed:`. When anthropic_content_blocks contained
only a blank cache-marked text block, `replayed` came out empty, the
function fell through to the main path's placeholder, and the cache
marker was lost; signed thinking + a blank marked text block likewise
returned with no cacheable carrier for the relocated marker.
The replay branch now appends the non-whitespace "(empty)" placeholder
when no cacheable (text/tool_use) block survives the blank filter and a
blank text block was dropped (or a marker needs a carrier) — so replay
stays schema-valid on Bedrock/strict endpoints and the breakpoint
survives on the placeholder.
Also reconciles the block-level tests from #69517 with the new
drop-then-fallback contract (blank blocks are dropped at the block
level; the message-level result is still always non-blank).
Refs #69512
Co-authored-by: ygd58 <buraysandro9@gmail.com>
Follow-up per independent review of #68633 (GPT-5.6-sol-xhigh in Codex,
reviewer egilewski) on this PR.
Two real bugs in the blank-text-block filtering added by that fix:
1. `effective = blocks or content` fell back to the RAW, unfiltered
`content` variable whenever every block was filtered out as blank --
which happens precisely when the entire message content WAS the
blank/whitespace payload the filter exists to remove (a sole blank
text block, a sole cache-marked blank block, or standalone
whitespace scalar content with no tool_calls). The fallback silently
restored the exact invalid content the filtering just stripped,
leaving the message provider-invalid.
Fixed: `effective = blocks if blocks else [{"type": "text", "text":
"(empty)"}]` -- never falls back to raw `content`. Also moved the
cache_control application (both the relocated-from-a-dropped-block
marker and the message-level marker) to run against `effective`
instead of the pre-fallback `blocks`, so a cache marker on a block
that was the ONLY content still lands on the (empty) placeholder
rather than being silently lost when `blocks` was empty at the
point it would otherwise have been applied.
2. The normal-path blank-text check used `(blk.get("text") or "").strip()`,
which is not type-safe for a truthy NON-string, non-None text value
(e.g. an int or dict from an invalid upstream payload) -- `or`
doesn't substitute for a truthy value, so `(7 or "").strip()` still
raises AttributeError. Now checks `isinstance(text, str)` first,
matching the replay path's `_sanitize_replay_block()`, which the
reviewer confirmed was already correctly type-safe.
Added regression tests for: sole blank list block, sole whitespace
scalar content, sole cache-marked blank block (marker relocation to
the placeholder), a truthy non-string (int) text value both mixed with
a surviving tool_use and as the sole content, and a dict-valued text
field. 7/7 new tests pass; 193/193 in the full
tests/agent/test_anthropic_adapter.py file; 23/23 in
tests/agent/test_prompt_caching.py (unaffected, confirmed).
Ports #63228 forward onto current main per teknium1's review.
Bedrock and strict Anthropic-compatible endpoints reject text blocks
where text is empty or whitespace-only with HTTP 400. The normal
list-content path extended blocks without filtering, and the
ordered-replay fast path (_sanitize_replay_block) returned blank text
blocks unfiltered.
Per review, fixes three gaps in the original port:
1. Type safety: the normal-path filter used blk.get('text', '').strip(),
which crashes with AttributeError when text is explicitly None (not
absent) -- .get()'s default only applies when the key is missing.
_convert_content_part_to_anthropic() can preserve None from an
invalid upstream input text block. Now uses
(blk.get('text') or '').strip() on both paths.
2. Cache marker loss: prompt_caching.py's _apply_cache_marker() sets
cache_control directly on content[-1] for list content. If that last
part happens to be blank text, dropping it without relocating
cache_control silently loses the breakpoint. Both the normal and
replay paths now capture a dropped block's cache_control and reapply
it to the new last surviving cacheable block via the existing
_apply_assistant_cache_control_to_last_cacheable_block() helper
(setdefault semantics, so it never clobbers a legitimately-placed
marker).
3. Scalar whitespace: the non-list content branch
(blocks.append({'type': 'text', 'text': str(content)})) accepted a
truthy whitespace-only string unfiltered. Now filtered the same way
as list-content blocks.
8/8 new tests pass in TestBlankTextBlockFiltering (including None-safety,
scalar-whitespace, and cache_control-relocation regressions on both
paths); 186/186 in the full tests/agent/test_anthropic_adapter.py file.
OpenAI documents GPT-5.5 / GPT-5.5 Pro as extended-cache-only: in-memory
prompt cache retention is not available for them, and only
prompt_cache_retention: "24h" is supported. Responses requests that omit
the field see near-zero cached_tokens even with a stable prompt_cache_key
and identical prefixes (observed on an OpenAI-compatible Responses relay:
0 cached across repeated identical calls before; 97% cache reads after).
Send the field for the gpt-5.5 model family (bare and namespaced ids like
openai.gpt-5.5) on OpenAI-compatible Responses routes, mirrored in the
auxiliary Codex adapter, and pass it through preflight normalization.
Skipped for xAI, GitHub/Copilot, and the chatgpt.com Codex backend, which
reject or ignore body-level cache fields.
A brew Python upgrade (original report) or an interrupted venv rebuild
(v0.19.0 report in the same thread) can leave certifi importable while
its bundled cacert.pem is missing or a dangling symlink. Every TLS
connection then fails — Feishu/Telegram/WeChat/DingTalk all down —
with an opaque 'Could not find a suitable TLS CA certificate bundle'
from deep inside httpx/requests.
The existing repair infrastructure only probed
`hasattr(certifi, 'contents')`, which PASSES in exactly this failure
state, so neither the early venv self-heal nor `hermes update`'s
import-probe repair ever classified certifi as broken. Extended, not
replaced:
- hermes_cli/_early_recovery.py: the in-process probe now also
validates that certifi.where() exists and is a plausible bundle
(>=1KiB), so the pre-import self-heal repairs it like any other
wiped core package.
- hermes_cli/main.py (_detect_broken_lazy_refresh_imports): the
subprocess probe script used by `hermes update`'s venv repair
applies the same bundle-file check inside the target venv.
- hermes_cli/doctor.py: `hermes doctor` already failed the cert
check; `hermes doctor --fix` now repairs it (pip force-reinstall
certifi + module-cache invalidation + re-verify), covering
brew/manual venvs where no update marker exists. Failures funnel
into the manual-action list with the exact command.
- agent/ssl_guard.py: the startup SSLConfigurationError hint now leads
with `hermes doctor --fix` instead of only the raw pip command.
Fixes#29866
Running any state-writing `hermes cron` CLI command as root (the
default for `docker exec`) rewrote jobs.json via mkstemp +
atomic_replace, leaving it root:root mode 600. The gateway's ticker
(uid 1000 via PUID/PGID) was then locked out of every tick with
PermissionError — silently: the liveness heartbeat stayed fresh,
`hermes cron status` opened with 'Gateway is running — cron jobs will
fire automatically', and in the field ~14h of scheduled jobs were
skipped before a human noticed the absence of messages.
Fixes, per the issue's suggested items 1 and 3:
1. Ownership preservation on save (cron/jobs.py): snapshot the owner
before the atomic replace; when the writer is privileged (euid 0)
and the previous owner differs, chown the rewritten file back.
First-time creation inherits the cron dir's owner. Unprivileged
writers never call chown. POSIX-only (guarded via os.name/getattr),
best-effort — a chown failure logs a warning but never breaks the
save. 0600 hardening is unchanged.
2. Zombie-ticker surfacing: the ticker loop (both single-profile and
multiplex paths) now persists the failure reason to a
ticker_last_error marker next to the heartbeat files on every
failed tick, and clears it on the next clean tick. `hermes cron
status` shows the recorded reason in its 'ticks may be failing'
branch, plus an actionable ownership hint when the error is a
PermissionError (recommend `docker exec -u <uid>:<gid>`).
Fixes#68483
Adds a 'Host integration' section to the ACP guide and a row in the
environment variable reference so the next ACP host implementer does not
have to read the adapter source.
Documents the exact contract the tests already pin: the value must be
exactly `1`; unset/empty/`0`/`false` keep the default behavior; only
globally configured config.yaml MCP discovery is skipped, and servers
supplied by the ACP session through session/new are still registered.
Framed as a host-set process marker rather than user configuration - the
same shape as the existing HERMES_KANBAN_TASK entry - so it does not read
as a behavioral setting that belongs in config.yaml.
Co-authored-by: amanning3390 <adam.manning@pro-serveinc.com>
Signed-off-by: amanning3390 <adam.manning@pro-serveinc.com>
ACP clients (Zed, Buzz) render the whole availableModels array in a single
dropdown, so requesting the shared inventory with max_models=None could
hand an editor an unbounded cross-provider catalog.
Request the same per-provider cap the MoA picker already uses
(hermes_cli/moa_cmd.py), exposed as ACP_MAX_MODELS_PER_PROVIDER so the
intent is documented at the call site.
This bounds each provider's row rather than the total, matching the shared
inventory's own semantics: aggregator providers stay intentionally
uncapped, and the existing current-model fallback still re-inserts a
selection that falls outside the cap. At present no authenticated provider
approaches 200 models, so the visible catalog is unchanged; the cap is a
guardrail for large catalogs (e.g. OpenRouter) rather than a change to
today's lists.
The new test asserts the contract - bounded row plus a reachable current
selection - instead of a fixed catalog size, so growing the inventory
cannot turn it into a change-detector.
Co-authored-by: amanning3390 <adam.manning@pro-serveinc.com>
Signed-off-by: amanning3390 <adam.manning@pro-serveinc.com>
When api_key_hint from a 401 response doesn't match any pool entry
(common with OAuth tokens where runtime_api_key rotates), the pool
rotated without marking anything exhausted and handed back a fresh
selection. Because nothing was ever marked, the pool could never reach
the "no available entries" state — the caller retried the same dead
token forever (~6 attempts/sec), starving the event loop so /stop was
never processed; only killing the gateway ended it.
Rebased onto the identity-tracking rework that landed on main
(73c4b5a045): the single-entry escape from that commit already stops
the most common OAuth case, so this fix bounds the REMAINING gap —
multi-entry pools ping-ponging A->B->A with an unmatched hint. Cap
consecutive no-mark rotations at one full lap of the available
entries, then return None so the error surfaces / fallback activates.
Deliberately does NOT mark innocent entries exhausted (the original
PR's approach): that would quarantine a healthy key for the full
cooldown TTL on a hint that provably matches nothing. No cooldown is
written by the escape, so healthy keys stay available next turn --
bounded without hammering.
The streak resets when a rotation identifies a real entry and on any
successful normal select(), so only genuinely consecutive unmatched
rotations trip the bound.
Fixes#70401
Follow-up on the #69500 salvage: 'launchctl submit' jobs remain
registered with launchd after they exit, so every plist reload leaked
one dead '<label>.reload.<pid>.<ts>' label. The helper script now ends
with 'launchctl remove' of its own transient label, and the recovery
test asserts the self-removal is present.
The deferred launchd reload helper used start_new_session=True to detach
from the gateway's process group. However, setsid(2) alone does NOT move
the child outside the launchd job's process coalition — when launchctl
bootout fires on the gateway label, launchd terminates ALL processes in
that coalition, including the setsid-detached helper, leaving the service
permanently unloaded.
Fix by spawning the helper via launchctl submit, which creates a
transient launchd one-shot job that is wholly independent of the
gateway's coalition. This ensures the helper survives bootout and can
complete the bootstrap+verify cycle.
Also writes a durable pre-bootout marker to the reload log so the
distinction between 'helper never started' and 'helper ran but
bootout/bootstrap failed' can be diagnosed.
Fixes#69098
Hardening follow-up to the #69619 review fix. The previous regression
byte-pinned only the rescued pre-#69619 generation; older frozen entries
were covered solely by fragment assertions and a self-matching loop that
cannot detect a frozen entry mutating (the loop tests each entry against
itself).
- Pin all four _HISTORICAL_SUMMARY_PREFIXES generations as literals in
_FROZEN_PREFIX_GENERATIONS and assert order-sensitive tuple equality
plus detect/strip for each
- State the prepend-only contract explicitly on the tuple: never mutate
or reorder existing entries
Negative controls verified: mutating, dropping, or reordering a frozen
entry each fail the new test, while the legacy self-matching loop still
passes under mutation — confirming the closed coverage gap.
Address review on #69619: the previous commit mutated the newest frozen
entry in _HISTORICAL_SUMMARY_PREFIXES and never froze the live prefix it
retired (the generation with both the four-heading discard clause and
the tools-active clause). A summary persisted immediately before
upgrading was therefore treated as an ordinary message on
resume/re-compaction, keeping the old handoff text embedded in the body.
- Prepend the exact pre-change live prefix as a new frozen entry
(newest-first), leaving all existing frozen entries byte-identical
- Restore the Jul 2026 (#65848 class) frozen entry to its original
four-heading text
- Pin the retired generation as a literal in
test_summary_prefix_semantics.py so mutating or dropping it fails CI
- Make the #65848 tool-use regression position-agnostic (match the
pre-clause generation by content, not tuple index)
Verified byte-identity of both rescued generations against the parent
commit. 233 focused prefix/resume/compressor tests pass.
Remove three directive-heavy section headers from both the LLM
and deterministic summary templates that caused the agent to
resume stale tasks after context compression:
- Historical In-Progress State
- Historical Pending User Asks
- Historical Remaining Work
These sections read as actionable instructions even within a
REFERENCE-ONLY wrapper, hijacking the user's latest message.
The remaining sections are purely descriptive/past-tense.
Frozen prefix copies in _HISTORICAL_SUMMARY_PREFIXES updated
to match. Test 8/8 passed.
Two gaps found auditing the decode-crash cluster:
1. suppress_platform_ver_console() only ran in hermes_cli.main processes;
slash workers, tui_gateway/entry, run_agent, batch_runner, and cli.py
import only hermes_bootstrap and were exposed to both the console
flash and (on Python 3.11.0/3.11.1, which lack CPython's
encoding='locale' fix) a UnicodeDecodeError inside platform.win32_ver()
under PEP 540 — the crash #69413 reported. Move the stub into
hermes_bootstrap so every entry point gets it; the _subprocess_compat
copy stays for non-bootstrap callers.
2. The desktop Electron spawn built the backend env without PYTHONUTF8,
so anything the Python child emitted before hermes_bootstrap ran
(interpreter startup errors, pre-bootstrap tracebacks) decoded with
the locale default. Re-port of PR #56499's env half (echoriver89) to
backend-env.ts (original targeted the deleted backend-env.cjs);
explicit user setting wins.
Salvaged from PR #45099 — the two popen_kwargs dict sites the #70875
AST sweep missed because the kwargs are built indirectly
(_run_command_stt, _run_command_tts).
Follow-up to the salvaged #38985: guard the 4 bare read_text/write_text
sites its allowlist missed (google_chat thread-count store + oauth JSON)
and add whatsapp/google_chat to the AST guard test's file list.
On Windows, pipe I/O can deliver non-UTF-8 bytes at chunk boundaries,
causing `UnicodeDecodeError` when the MCP SDK's `TextReceiveStream`
uses `errors="strict"`. Set `encoding_error_handler="replace"` on
`StdioServerParameters` so undecodable bytes become U+FFFD instead
of crashing.
Add a persisted, backend-confirmed provider/model lock for Hermes
Browser and other session API clients. A confirmed lock is an
execution contract rather than response metadata:
- POST /api/sessions/{session_id}/model validates and persists a
confirmed browser_model_lock (advertised in /v1/capabilities)
- session chat + chat/stream consume the persisted lock on body-only
follow-up turns; a confirmed lock wins over an older gateway session
/model override and the session-persisted model
- a later successful session /model switch explicitly clears and
replaces the lock while preserving lineage markers (_branched_from)
and invalidating cached system-prompt model/provider metadata
- ordinary one-off request overrides never replace a confirmed lock
- provider-resolution failure fails closed as a typed provider-auth
error (controlled response, never global-credential reuse)
- confirmed locks disable the global fallback model chain
- the completed agent's actual provider/model must match the locked
route or the turn fails with a runtime-mismatch error
- responses carry sanitized runtime metadata reporting actual vs
requested provider/model and lock state
Rebased onto the provider-aware request routing (#70853) and
session-model parity (#70931) that landed since the original branch;
the lock now slots into that precedence chain as the top rung.
Salvaged from PR #61236 by @abundantbeing.
Anthropic released Claude Opus 5 (+ -fast variant) — both are live on
OpenRouter and the Nous Portal /models endpoint (verified against both
live APIs). Opus 4.8 entries are kept.
- hermes_cli/models.py: opus-5 + opus-5-fast in OPENROUTER_MODELS;
opus-5 in _PROVIDER_MODELS[nous] (Portal serves both, curated list
carries the base model like the rest of the Nous Anthropic block).
Ordering: below fable-5 flagship, above opus-4.8.
- agent/model_metadata.py: claude-opus-5 -> 1M context (matches live
OpenRouter metadata).
- agent/reasoning_timeouts.py: claude-opus-5 -> 240s stale-timeout
floor (same as the opus-4.x thinking family).
- website/static/api/model-catalog.json: regenerated via
scripts/build_model_catalog.py.
Both providers bill via official_models_api (live pricing), so no
_OFFICIAL_DOCS_PRICING snapshot entry is needed for these routes.
Three parity fixes between the API server and the native gateway's
agent-runtime resolution, integrated with the provider-aware request
routing that landed in #70853:
- Session-persisted model is honored: POST /api/sessions {"model": ...}
stores a model that the chat handlers previously fetched and threw
away. A stored value that matches a model_routes alias goes through
the route path (route provider/credentials apply); a raw model string
threads through as session_model, pinning the session's turns ahead
of per-request body values but below an explicit session /model
override.
- Empty-model recovery: provider-catalog default when config has no
model.default but a provider resolved, plus last-known-good model
recovery (#35314) keyed on gateway_session_key only (never ephemeral
session_id — no unbounded growth from one-off requests).
- Provider auth failures surface as controlled responses: RuntimeError
from _resolve_runtime_agent_kwargs() is re-raised as a dedicated
_ProviderAuthResolutionError at the call site, caught narrowly in
_run_agent() and the /v1/runs executor to return run.py's response
shape instead of an undifferentiated 500 (session-chat endpoints
previously returned a raw aiohttp 500 with no JSON body).
Salvaged from PR #57947 by @FvanW; session-model route-alias resolution
from PR #59941 by @kaishi00.
Co-authored-by: kaishi00 <kaishi00@users.noreply.github.com>
cron/scheduler.py deliberately applies utf-8/replace only on Windows via
popen_kwargs (non-Windows keeps locale default per its test contract) —
drop the sweep's unconditional inline kwargs there. Update the gateway
force-kill kwarg snapshot for the new guard.
- whatsapp taskkill + webhook gh-comment assert_called_with: add the two
new kwargs
- test_status fake_run: accept **kwargs so signature-strict stub doesn't
TypeError on encoding/errors
- Strip the salvaged commit's inline encoding kwargs where main had since
gained its own (process_registry, local env, cua doctor, gateway,
commands, gateway_windows — the latter keeps its locale-aware
_schtasks_encoding() from #38186)
- Revert encoding kwargs mistakenly applied to non-subprocess APIs
(exa get_contents, tempfile.mkstemp in webhook.py)
- Guard the ddgs worker Popen (new on main since #55339)
- Update two kwarg-snapshot test assertions for the new kwargs
AST-driven pass over every subprocess.run/Popen/check_output/check_call/call
with text=True (or universal_newlines=True) and no explicit encoding=:
append encoding='utf-8', errors='replace' at the kwarg site. 136 call
sites across 28 files (cli.py, hermes_cli/main.py, tools_config.py,
environments, computer_use, gateway, scripts, skills helpers, agent/*).
Together with the salvaged #55339/#60741 commits this closes out issue
#53428's bug class; the salvaged #60751 linter rule in
check-windows-footguns.py now enforces it repo-wide (verified: 807 files
scanned, zero findings).
Adds a new rule to scripts/check-windows-footguns.py that flags
subprocess.run/Popen/call/check_output/check_call(..., text=True, ...) calls
missing an explicit encoding= kwarg.
On Chinese Windows (cp936/GBK) and other non-UTF-8 default codepages,
text=True without encoding= decodes child output with
locale.getpreferredencoding(False), crashing _readerthread with
UnicodeDecodeError on non-default-codepage bytes (issues #47939, #53428,
rule prevents future regressions.
Rule design:
- Pattern matches 'text=True' / 'text = True'
- post_filter skips lines that:
- already pass encoding= on the same line
- are method definitions (def text)
- contain text=True inside string literals
- are not subprocess-shaped calls (heuristic via _is_likely_subprocess_call)
- Two helper functions: _is_likely_subprocess_call, _looks_like_string_literal
- Multi-line calls where subprocess.X( and text=True are on different lines
are not flagged (acceptable false negative for a line-based scanner)
Also fixes the linter's own footgun: get_staged_files() and get_diff_files()
used subprocess.check_output(text=True) without encoding= — now fixed.
Suppresses 4 false positives on non-Windows platform-exclusive calls:
- tools/voice_mode.py (Termux/Android)
- tools/environments/singularity.py (Linux HPC)
- plugins/google_meet/cli.py (macOS system_profiler)
Test plan:
- 21 unit tests in tests/scripts/test_footgun_subprocess_encoding.py
- TestDetection: 6 cases verifying the rule flags real subprocess calls
- TestSuppression: 7 cases verifying false-positive avoidance
- TestHelpers: 7 cases for the two helper functions
- TestFullRepoScan: scans the whole tree and asserts the new rule finds
only the 7 call sites that PR #60741 fixes (or zero, once #60741 merges)
Verified: full-repo scan reports 7 matches on main (the #60741 sites),
4 platform-exclusive calls correctly suppressed, zero false positives.
Hermes-sweeper review on #60741 flagged that _op_version (the paired
op probe used by the same setup/status CLI flow at lines 127 and 205)
still ran text=True without explicit encoding/errors.
Add encoding='utf-8', errors='replace' to match _op_whoami and the
production op read path at agent/secret_sources/onepassword.py:271-278.
Also extend the regression test to cover _op_version alongside
_op_whoami, and update the module docstring to reflect the widened
scope. Test sensitivity verified: reverting the source change makes
test_op_version_passes_utf8_encoding fail with encoding=None.
PR #55339 adds encoding='utf-8', errors='replace' to 26 subprocess.run(text=True)
call sites across the codebase. The triage review (thanks @alt-glitch) diffed
this PR against #55339 and found that 5 of the 6 originally-touched call sites
are already covered there byte-identically:
- hermes_cli/main.py::_probe_container
- hermes_cli/setup.py SSH probe
- tools/tts_tool.py::_generate_neutts
- tools/transcription_tools.py::_prepare_local_audio
- tools/transcription_tools.py::_transcribe_local_command (both branches)
The one genuinely net-new site — hermes_cli/onepassword_secrets_cli.py::_op_whoami
(the 1Password op CLI whoami probe) — is NOT in #55339 and is fixed here.
Without explicit encoding=, text=True decodes child output with
locale.getpreferredencoding(False) — cp936 on Chinese Windows — which crashes
_readerthread on non-GBK bytes, cascading into pipe buffer fills, event loop
stalls, and TUI freezes (issues #47939, #53428, #57238).
Scope narrowed per triage feedback: the other 5 sites should land via #55339.
Refs #53428 (together with #55339).
On Windows with Chinese locale (GBK), subprocess.run(text=True) without
explicit encoding causes UnicodeDecodeError crashes. This fix adds
encoding='utf-8', errors='replace' to all subprocess.run() and
subprocess.Popen() calls that use text=True across 76 non-test Python files.
Fixes#53428 (master tracker for Windows GBK locale crash).
Note: credential_pool.py and electron changes excluded per reviewer request —
those will be submitted as separate focused PRs.
* feat(desktop): add Cron Blueprints to the GUI
The desktop app had a Cron jobs panel but no Blueprints tab, so the
parameterized automation templates the dashboard offers were unreachable
there (parity matrix: Dashboard=Y, GUI=N).
Adds a Jobs/Blueprints segmented toggle to the cron panel. The Blueprints
tab renders the catalog from the existing GET /api/cron/blueprints endpoint,
one card per template with a typed form (time/enum/weekdays/text slots).
Submitting POSTs to /api/cron/blueprints/instantiate, which fills the slots
and creates a real cron job via the same create_job path as a hand-written
one. The new job is merged into the shared atom so the Jobs tab
and sidebar reflect it immediately.
Backend already served both endpoints; this is desktop frontend only.
Strings added to all four locales.
* fix(desktop): stop cron blueprint cards clipping and tabs overlapping close X
Blueprint cards were wrapped in PanelBlock (a max-h-48 overflow-auto <pre>
for monospace code), which capped each card and forced an inner scroll,
clipping the copy. Use a plain auto-height card div so rows grow with their
content and the gallery scrolls as one.
PanelHeader actions sat under the overlay's absolutely-positioned close X
(no layout space reserved). Reserve pr-8 clearance when actions are present.
* refactor(desktop): narrow blueprint card i18n dep, document intentional scoping
Address review nits on the Cron Blueprints PR:
- BlueprintCard's submit useCallback depended on the whole t.cron object; it
only uses the blueprints slice. Bind const b = c.blueprints and use it (plus
narrow the dep) throughout the card.
- Document the intentional GET-vs-POST profile asymmetry on the blueprint
endpoints (global catalog vs per-profile instantiate) — the prior comment
claimed both were profile-scoped.
- Note why the blueprints tab collapses 'all' scope to 'default' (a blueprint
creates a real per-profile job; 'all' is not a writable target).
* fix(desktop): default blueprint delivery to This desktop, not origin
The blueprint catalog is shared with the dashboard, so its deliver slot
defaults to 'origin' (the chat/home-channel a dashboard or gateway job was
created from). Desktop has no origin chat and no home-channel picker, so the
seeded 'origin' rendered unlabeled and, at delivery, fell through the
home-channel fallback to nowhere when no gateway was configured.
Seed the deliver slot to 'local' (This desktop) when the backend default is
'origin' or empty, drop the origin option from the desktop dropdown, and label
the remaining options with the desktop's own delivery labels — matching the
manual cron editor (local/telegram/discord/slack/email). Also skip the
backend's origin-centric deliver help, which contradicts desktop semantics.
* refactor(desktop): align blueprint cards with the Panel/settings idiom
Address PR review (UI consistency with neighboring surfaces):
- Card container: drop the standalone-card look (bg-foreground/5) for the
shared in-panel grouping token bg-(--ui-bg-quinary), matching the cron
editor's in-surface groupings so blueprints sit in the Panel family.
- Form fields: replace the bespoke <label>+<Input> rows with the shared
ListRow primitive from settings/primitives (label+help on the left, control
on the right, stacks in a narrow pane) — same idiom as settings/messaging.
No behavior change; blueprint deliver remap and $cronJobs merge untouched.
* fix(desktop): satisfy eslint on the cron blueprint files
CI check:lint failed on import/export ordering and an unused import in the
blueprint changes:
- hermes.ts: sort AutomationBlueprint before AuxiliaryModelsResponse in the
type import + re-export blocks, and drop the unused AutomationBlueprintField
import (still re-exported for blueprints.tsx).
- cron/index.tsx: alphabetize the dialog/segmented-control imports and the
./blueprints vs ../shell/statusbar-controls group.
- blueprints.tsx: add the required blank lines between statements.
eslint --fix only; no behavior change. typecheck + blueprint tests green.
* refactor(desktop): blueprints reuse the cron editor dialog + shared card
Address review: the blueprint UI was still going its own way on the card and
form. Reuse the app's canonical pieces instead of a bespoke surface.
- CronEditorDialog gains a 'blueprint' mode: EditorState carries the blueprint
+ target profile, the dialog renders the typed slots with the same
Field/FieldHint/DialogFooter/error-block chrome as manual New cron, and
submit routes to instantiateAutomationBlueprint. One dialog, one editor state
machine. Resolves the accordion, the border-t divider, ListRow-vs-Field, the
ad-hoc buttons, and the plain error <p> in one move.
- Blueprint cards use selectableCardClass({ prominent: true }) (the shared
theme/pet/gateway card idiom), caller owns padding (p-2), whole card is a
button that opens the dialog pre-filled. No inline form.
- Gallery renders via PanelDetail, not PanelBody's master/detail row.
- i18n: drop the now-unused blueprints.setUp/cancel, add blueprints.dialogDesc
across en/ja/zh/zh-hant + types.
- Dropped the stray \u2014 literal comment.
Logic (origin->local deliver, desktopDeliverOptions, merge into $cronJobs) is
unchanged and stays unit-tested. typecheck + eslint + tests green.
* fix(desktop): dropdown no longer closes the cron dialog; unify deliver targets
Two cron-dialog bugs:
1. Dismissing any Select dropdown inside the cron editor dialog closed the whole
dialog. Radix portals Select/Popover content outside the dialog, so the
dismiss pointerdown reached the Dialog's DismissableLayer as an
outside-interaction. Guard DialogContent.onInteractOutside: swallow
interactions originating from a [data-radix-popper-content-wrapper] (a
dropdown dismiss inside our own dialog), compose with any caller handler. Fix
is at the shared Dialog level so every dialog benefits.
2. Blueprint deliver only offered 'This desktop'. The blueprint used the backend
blueprint field.options (configured gateways only) while the manual editor
hardcoded local/telegram/discord/slack/email regardless of what's connected.
Wire the desktop to GET /api/cron/delivery-targets (the documented single
source of truth, already used by the dashboard) via getCronDeliveryTargets,
and render both the manual editor and the blueprint deliver slot through one
shared DeliverSelect. Now all three surfaces agree and only offer connected
platforms; unconfigured-home-channel targets show a hint.
i18n: add cron.deliverNeedsHomeChannel across en/ja/zh/zh-hant + types.
typecheck + eslint + vitest green.
* fix(desktop): clicking away from an open dropdown no longer closes the dialog
The onInteractOutside guard only caught pointerdowns whose target was inside
the popper wrapper. But dismissing an open Select by clicking elsewhere inside
the dialog also closes the popover, which moves focus — and Radix Dialog reads
that as focusOutside and closes the whole dialog. (Radix Select 2.3.1 has no
modal prop, so that escape hatch isn't available.)
Guard both paths at the shared DialogContent level: onInteractOutside AND
onFocusOutside now swallow the event when it originates from a Radix popper OR
when any [data-radix-popper-content-wrapper] is open at event time (covers the
focus/re-dispatch case where the target is no longer the popper). A genuine
backdrop click with no dropdown open still closes the dialog. Export
isInteractionFromPopper + unit-test the three cases.
typecheck + eslint + vitest green (7 dialog tests).
* fix(desktop): portal popovers into their dialog so dropdowns don't close it
Root cause (affected every dialog, not just cron): Radix Select/Popover/
DropdownMenu portal to document.body — a SIBLING of the dialog, outside its DOM
subtree. Dismissing a dropdown (or clicking another field) moves focus out of
the dialog subtree, which the Dialog's modal FocusScope/DismissableLayer reads
as an outside interaction and closes the whole dialog. Separate body-level
portals also make z-index across the two fragile.
The earlier onInteractOutside/onFocusOutside guards treated symptoms and didn't
hold (and Radix Select 2.3.1 has no modal prop to disable its layer). Real fix
is a layering system: DialogContent publishes its content node via
DialogPortalContainerContext; SelectContent/PopoverContent/DropdownMenuContent
call usePopoverPortalContainer() and portal INTO that node when inside a dialog
(document.body otherwise). The popover is then a true DOM descendant of the
dialog — focus stays in, dismissal no longer closes the dialog, and both share
one stacking context so z-index is deterministic.
Test: with a Dialog open, an open Select's item is a descendant of the dialog
(portalled in), verified in jsdom. typecheck + eslint + component tests green.
* fix(desktop): bump radix-ui so dismissing a dropdown can't close its dialog
Upstream bug, not app-layer: with radix-ui 1.6.0 (dismissable-layer 1.1.13),
an open modal Select sets pointer-events: none on the dialog body, so a click
anywhere inside the dialog hit-tests through to the overlay. The Dialog's
DismissableLayer defers its outside-pointerdown decision to the click, but the
overlay is a registered dismissable surface exempt from the interception check
— so the Select swallowing the press didn't count, the deferred onDismiss
fired, and the dialog closed along with the dropdown.
dismissable-layer 1.1.17 adds shouldHandlePointerDownOutside, which makes the
dialog's layer ignore the press entirely while a higher layer has its pointer
events disabled. Bump radix-ui ^1.4.3 -> ^1.6.5 (dismissable-layer 1.1.17,
dialog 1.1.21, select 2.3.5); lockfile diff is Radix-only.
Repro test fires the pointerdown/up/click sequence on the overlay with a
Select open inside the dialog: red on 1.6.0 (dialog closed), green on 1.6.5.
Counterpart test keeps a genuine overlay click closing the dialog.
typecheck + dialog/select component tests green. Full-suite failures on this
Windows host reproduce identically without the bump (pre-existing env issues).
Add authenticated GET /api/model/options to the gateway API server,
sharing the dashboard/TUI picker payload builder so external clients
can sync to the user's configured Hermes provider catalog instead of
scraping the single OpenAI-compatible /v1/models alias.
- new shared hermes_cli.inventory.build_model_options_payload() wraps
build_models_payload with the stable picker shape and safe
custom-provider probe policy (probe current only on normal open,
probe all + cache bust on explicit refresh)
- dashboard web_server and TUI gateway model.options refactored onto
the shared builder; dashboard build moved off the event loop via
run_in_threadpool
- capabilities endpoint advertises model_options
- docs for both API server and programmatic integration
Salvaged from PR #54689 by @abundantbeing.
* feat(desktop): add Webhooks page for subscription CRUD
Brings the desktop GUI to parity with the dashboard's Webhooks page.
Adds a /webhooks route that lists webhook subscriptions, enables the
webhook gateway platform, and creates/toggles/deletes subscriptions,
hitting the same /api/webhooks* endpoints the dashboard and CLI use.
- types/hermes.ts: WebhookRoute, WebhooksResponse, WebhookCreatePayload,
WebhookCreateResponse, WebhookEnableResponse
- hermes.ts: getWebhooks, enableWebhooks, createWebhook, deleteWebhook,
setWebhookEnabled (profile-scoped) + type re-exports
- app/webhooks/index.tsx: WebhooksView (enable card, restart banner,
subscription list with copy/toggle/delete, create dialog with
one-time secret reveal); optimistic toggle, profile re-home
- routing: routes.ts, contrib/surfaces.tsx, chat/route-tile.tsx
- nav: command palette, keybinds (nav.webhooks), sidebar row
- i18n: en + zh full, types interface; ja/zh-hant fall back to English
- test: webhooks-rest.test.ts covers the REST helper contracts
* feat(desktop): surface Webhooks in the status bar instead of nav
Moves the Webhooks entry point from the sidebar nav / command palette /
keybind to a status bar action next to Cron, matching where scheduled
jobs live. The /webhooks route, page, and REST helpers are unchanged.
- use-statusbar-items.tsx: add webhooks action (Globe icon) after cron
- i18n: shell.statusbar.webhooks / openWebhooks (en, zh, types)
- revert nav wiring: sidebar row, command palette entry, nav.webhooks
keybind action + label, commandCenter.nav webhooks entry
* feat(desktop): add skills field to webhook create form
Exposes the backend's per-subscription skills list in the create dialog
(comma-separated) and shows skill badges on subscription rows, so this
page covers the skill-backed endpoint case as well as general CRUD.
- create form: Skills input; passes skills[] to createWebhook
- rows: render skill badges
- i18n: fieldSkills / fieldSkillsPlaceholder (en, zh, types)
Consolidates the skill-endpoint framing from #42817.
Co-authored-by: LionGateOS <98371158+LionGateOS@users.noreply.github.com>
* feat(desktop): render Webhooks as an overlay instead of a full page
Webhooks took over the whole workspace/chat pane because 'webhooks' was
missing from OVERLAY_VIEWS, so it routed through PageSearchShell while
cron rendered through the Panel overlay. Add it to OVERLAY_VIEWS and wire
it up like cron: mount WebhooksView as a floating overlay in wiring.tsx,
render null for the webhooks route in the workspace table, and convert
WebhooksView from PageSearchShell to the Panel primitive with an onClose.
Drop webhooks from route-tile BUILTIN_PAGES so it can't be tiled as a
page either.
* fix(desktop): place webhook URL copy button next to the URL
The URL span used flex-1, stretching it across the row and pushing the
copy button to the far right. Drop flex-1 so the span sizes to content
and the copy button sits directly after the URL.
* fix(desktop): top-align the deliver-only checkbox with its wrapped label
The label used h-9 items-center, centering the checkbox against the
two-line hint text. Switch to items-start so the checkbox aligns to the
first line.
* feat(desktop): give Webhooks a cron-style master/detail layout
Replace the single-column subscription list with the same Panel
master/detail cron uses: a left PanelList of subscription rows (status
dot + kebab menu) capped by a PanelAddButton, and a right PanelDetail
showing the selected subscription's deliver/events/skills, URL with copy,
description, and prompt. Enable/restart banners sit above the body; the
empty state keeps its own New subscription action.
* fix(desktop): drop header on the Webhooks empty state to match cron
The zero-subscriptions state still rendered the PanelHeader with the
title and the refresh/new buttons on the right. Remove it so the empty
state is just the centered PanelEmpty (icon, message, New subscription),
matching the cron empty state. Also drop the header from the loading
state.
* fix(desktop): top-align deliver-only checkbox and its wrapped label
Drop the min-h-9/pt-1.5 baseline shim that pushed the row down and made
the wrapped hint look misaligned. Use plain items-start with a mt-0.5 on
the checkbox so it sits at the first line, and wrap the hint in a
leading-snug span.
* fix(desktop): drop header refresh/new buttons; + button owns create flow
The populated Webhooks header carried a refresh icon and a New
subscription button. Remove both — the PanelAddButton at the bottom of
the list is the create flow, matching cron. Profile-change reload and the
refresh hotkey still run; the restart banner keeps its own refresh.
* fix(desktop): nudge deliver-only checkbox down 2px
Bump the checkbox top margin from mt-0.5 to mt-[4px] so it sits level
with the first line of the wrapped hint.
* refactor(desktop): reuse shared primitives on the Webhooks page
Address OutThisLife's review — stop reinventing primitives the app
already ships:
- copy: drop the local navigator.clipboard button for the shared
CopyButton (routes through the Electron clipboard bridge + haptic +
error state instead of swallowing failures)
- banners: enable/restart callouts now use Alert variant=warning
(primary color-mix tokens) instead of a hand-rolled amber palette
- toggle: detail Enable/Disable is a Switch (messaging idiom), not a
text ghost button
- checkbox: create dialog uses the Checkbox primitive, not a raw input
Rows/chips already moved to PanelListRow/PanelDetail/PanelPill in the
earlier cron-layout pass. Left the main-list fetch on manual load and
the local Field helper: cron itself does both, so useQuery here would
diverge from the reference idiom rather than align with it.
* refactor(desktop): move Webhooks fetch to the react-query layer
Replace the manual useState/useEffect load with useQuery keyed by
['webhooks', profileScope] — profile change re-fetches automatically, no
effect. reload() invalidates the query; the optimistic toggle writes the
cache via queryClient.setQueryData then invalidates so backend truth
wins. Load failures surface via an error-watching effect (react-query v5
dropped useQuery onError). Refresh hotkey calls refetch().
Left the create-dialog Field helper as-is: settings ListRow is a
side-by-side settings row, wrong for a stacked dialog form, and cron's
editor dialog (the reference) defines the same local Field.
* refactor(desktop): drop bespoke Field for shared ListRow/ToggleRow
Remove the local Field wrapper entirely. Every create-dialog field now
uses settings ListRow (wide, so label stacks over the full-width
control), the deliver-only pref uses ToggleRow (ListRow + Switch, haptic
baked in) instead of a bare Checkbox, and the created URL/secret reveal
rows use ListRow too. No component in this file is hand-rolled anymore.
* fix(desktop): pair webhook create fields into a 2-column layout
The single-column dialog scrolled awkwardly. Group fields: name +
description side by side, prompt full-width under them, events + skills
side by side, deliver-to + deliver-only side by side. Drop the
deliver-only help text and rename the label to 'Deliver payload only'
(remove the now-unused fieldDeliverOnlyHint i18n key from en/zh/types).
* fix(desktop): align Webhooks page with cron conventions and DESIGN.md
- Delete copy uses deleteDescPrefix + bolded name + deleteDescSuffix (no em-dash)
- Drop the duplicated detail-header Switch + Trash2; enable/disable and delete
live only in the row kebab, matching CronJobDetail
- Collapse three copyable-value chromes into one flat token-backed CopyValueRow
- Delete success toast gains a w.deleted title
- Replace hand-rolled delete Dialog with shared ConfirmDialog
---------
Co-authored-by: LionGateOS <98371158+LionGateOS@users.noreply.github.com>
Adds a first-run Desktop choice between installing Hermes locally and
connecting to an existing remote Hermes gateway. The choice appears after
backend resolution but before ensureRuntime(), so selecting remote cannot
accidentally trigger local bootstrap.
New modules:
- first-run-setup-gate: concurrent first-run decision gate and reset semantics
- primary-backend-startup: Electron-free orchestration seam (saved remote
resolution, gate decision, remote re-resolution, local continuation)
- primary-connection-rehome: prevents dual-owner race where both cold boot()
and renderer softSwitch() could connect simultaneously
- first-run-remote-form: extracted remote form with stale-result guards
Reuses existing connection-config IPC, encrypted token storage, OAuth
session partition, and primary backend resolution.
Fixes#38602Fixes#36970
Follow-up to salvaged PR #70549:
- Replace fragile 'or' assertions with single precise checks that catch
partial-rewrite regressions (would have masked a missing closing brace)
- Add test_pty_path_uses_rewritten_command covering the PTY spawn path
that was modified but previously untested
Base-url+model dedup was meant for custom shim aliases, but it also
skipped first-class providers that share an inference host while using
different credentials. That stranded xai-oauth spending-limit failover
to the xai API-key provider when both used the same model slug.
A session pinned to a named custom provider could silently reroute to the
user's default provider on resume/rebuild. Session rows persist the RESOLVED
provider — bare "custom" for every named providers:/custom_providers: entry —
and when no base_url survived in model_config, the existing heal
(canonical_custom_identity) had only the config.model.provider fallback left.
For users whose global default is a BUILT-IN provider (e.g. nous) that tier
cannot fire, so the bare provider was dropped, resume fell back to the default
provider with the session's custom model name, and the default endpoint 404'd
with "Model '<x>' not found. The requested model does not exist in our
configuration or OpenRouter catalog." Re-selecting via /model fixed it until
the next resume — the reported symptom.
Add a model-name recovery tier between the base_url reverse-lookup and the
config fallback: find_custom_provider_identity_by_model() maps the stored
model back to the entry that serves it (model/default_model/models catalog,
dict and legacy list shapes). The session row always stores the model, so the
entry identity survives even when the row has no base_url AND the global
default points elsewhere.
All five bare-custom heal sites in tui_gateway/server.py now pass the model:
_ensure_session_db_row, _stored_session_runtime_overrides,
_runtime_model_config, _make_agent, and _model_picker_context.
Stale truncate_before_user_ordinal=0 from a desynced Desktop client
resolved to history[:0] and replace_messages() wiped the durable
transcript. Require confirm_empty_truncate for that edge and have
intentional first-turn restore/regenerate paths send it.
The Telegram adapter's connect retry loop could silently stall after
'Connecting to Telegram (attempt 1/8)...' with the event loop permanently
parked in select() — all threads idle, no attempt 2/8 ever scheduled.
Root cause analysis:
- The retry loop reused the same Application object across all
8 attempts. After a failed initialize() the app could be in a partially-
initialized state (closed httpx transports from ,
or flag set before the hang) causing subsequent calls
to silently skip real initialization.
- CancelledError (a BaseException, not an Exception) propagated silently
through all except handlers with no logging — the task driving the retry
loop could exit without any trace.
- No total watchdog bound existed for the entire retry loop; only per-attempt
timeouts via _await_with_thread_deadline. If the loop itself stalled
between attempts (between-attempt sleep, cleanup, or scheduling), there
was no timeout to catch it.
Fixes:
1. **Total watchdog deadline**: Compute a total deadline for the entire
connect loop (8 attempts × init_timeout + 120s margin). Before each
attempt, check the wall clock; if exceeded, raise OSError immediately
instead of attempting another initialize().
2. **Fresh Application per retry**: On each failed attempt, rebuild
via and re-register all handlers. The old
app is best-effort shutdown with . This ensures
each retry starts with a clean slate — no stale transports, no stale
flag, no leaked state from the previous attempt.
3. **BaseException logging + propagation**: Added
(placed LAST after all other handlers) to log CancelledError and other
non-Exception signals before propagating. Previously these exited the
retry loop silently with no log message.
4. ** block for app rebuild**: The clause runs after
every failed attempt that isn't the last, rebuilding the app and
discarding the old one regardless of which exception class caused the
failure.
Issue #68915: when the agent runs a compound command with trailing & (e.g.
`cd /app && node server.js &`), bash parses it as `(A && B) &` — a subshell
that holds the stdout pipe open forever when B is a long-running server.
The existing _rewrite_compound_background in terminal_tool.py correctly
rewrites this to `A && { B & }` to avoid the subshell fork, but it was only
applied in the foreground execute() path (tools/environments/base.py).
The background spawn_local() path bypasses base.py entirely and passed the
raw command directly to Popen/PTY, leaving the deadlock unmitigated.
Fix: apply _rewrite_compound_background in spawn_local() before the command
is passed to Popen or PTY spawn. Uses a lazy import to avoid circular
dependency (terminal_tool imports process_registry).
- PTY spawn path: now uses safe_command (rewritten)
- Popen spawn path: now uses safe_command (rewritten)
- Session.command still stores the original (unrewritten) command for display
- Simple `cmd &` is left unchanged (no subshell bug)
Tests: 4 regression tests verifying (1) compound is rewritten, (2) simple bg
is preserved, (3) multi-line compounds are rewritten, (4) session.command
stores original.
Follow-up on the salvaged #68469 commit:
- Literal-IP hostnames never take the proxy DNS-delegation path (a
getaddrinfo failure on a literal IP is not a proxy-environment
symptom, and IPs need no DNS) — keeps the private-IP/metadata floor
intact under proxy env vars.
- Adds TestProxyEnvironmentDnsDelegation: delegation fires only for
hostnames, metadata hostname/IP floor holds, DNS-success path
unchanged, empty proxy var ignored.
- Guards the three pre-existing DNS-failure tests against ambient
proxy env vars so they don't flake on developer machines.
When the runtime blocks direct DNS (NVIDIA OpenShell, Docker + Squid,
corporate proxy with DNS-only-via-proxy), socket.getaddrinfo() fails
and is_safe_url() blocks *all* requests — including legitimate public
URLs via the configured proxy.
Add _proxy_is_configured() helper that checks HTTPS_PROXY, HTTP_PROXY,
http_proxy, https_proxy, ALL_PROXY, all_proxy. When DNS fails AND a
proxy is configured, delegate DNS resolution to the proxy rather than
blocking outright.
Blocked hostnames (metadata.google.internal, 169.254.169.254, etc.)
are checked BEFORE DNS resolution, so cloud metadata endpoints remain
blocked regardless of proxy status.
Fixes#32217
Replace 3 duplicated entry_id resolution blocks (try/except +
entry_id_for_api_key + fallback to None) in agent_init.py,
chat_completion_helpers.py, and switch_model with a single
sync_credential_pool_entry_id(agent) function in agent_runtime_helpers.
Follow-up to #70323.
Track the selected credential by stable pool entry ID so token refreshes and shared cursor movement cannot detach failures from the entry that issued them. Stop unmatched single-entry pools from reporting a no-op rotation as successful recovery.
Co-authored-by: Maxim Esipov <maksesipov@gmail.com>
Registers `ar` in the supported-language set and alias table and ships
locales/ar.yaml at full key and placeholder parity with en.yaml, covering
approval prompts and gateway slash-command replies. Identifiers, commands,
paths, config keys, model/provider names, and {placeholder} tokens are kept
verbatim.
Co-authored-by: Da7-Tech <286182457+Da7-Tech@users.noreply.github.com>
Adds the Arabic catalog to the dashboard, registers it in the locale list
and picker, and flips the document direction to RTL when Arabic is active.
Introduces a `defineLocale` merge helper (mirroring the desktop app) so the
Arabic catalog can be a partial override that falls back to English for any
untranslated key instead of hand-porting every future string.
Co-authored-by: morolab <ahmedmoro@gmail.com>
Arabic is the desktop app's first right-to-left locale. The i18n provider
now sets `document.dir`/`lang` from the active locale so Tailwind logical
utilities flip automatically, and `ar` is registered in the catalog,
language options, and alias table. The catalog is a partial `defineLocale`
so keys added to English later fall back cleanly.
Co-authored-by: 3ssiri <assiri@gmail.com>
Co-authored-by: Da7-Tech <286182457+Da7-Tech@users.noreply.github.com>
Follow-ups on the salvaged #54426 routing contract:
- Bare `model` without `provider` on the OpenAI-compatible endpoints
(/v1/chat/completions, /v1/responses) is now opt-in via
gateway.platforms.api_server.direct_model_requests (default off) —
generic OpenAI clients hardcode model names ('gpt-4o', ...) and
existing deployments rely on those falling back to the gateway
default. Explicit `provider` requests and the Hermes-native
session-chat + /v1/runs surfaces are always honored.
Idea credit: PR #22825 by @mssteuer.
- A model_routes alias with no `model` key can no longer leak the
alias string as the executing model name (defensive; parse-time
validation already drops such routes).
- Fix mis-indented _run_agent call args in _handle_session_chat_stream.
- Docs: document the opt-in flag.
Carry model, provider, and model_options through the API server's
execution surfaces (session chat, Chat Completions, Responses, /v1/runs)
without mutating global configuration. Precedence: session /model
override -> model_routes alias -> direct request selection -> global
defaults. Conflicting route/provider mixes fail closed with 400.
model_options stays request-scoped regardless of which selection wins.
Salvaged from PR #54426 by @abundantbeing.
test_docker_network_config.py landed on main after the #58489 revert and
stubbed docker ps with the 2-field ID\tState format. The re-landed
egress-aware reuse probe requests ID\tState\tEgressLabel when egress is
off, so the fake line failed to parse and the reuse path never fired.
Fixture-only change; production behavior is unchanged.
The salvaged #61978 covers tui_gateway/server.py. The crash reported on
Jul 24 came from a sibling site it doesn't touch: the desktop update
panel's _recent_upstream_commits() in hermes_cli/web_server.py runs
git log with text=True and no encoding. Commit 84db32484f put a bug
emoji (UTF-8 f0 9f 90 9b) in a subject on main; byte 0x90 is undefined
in cp1252, so every Windows desktop install behind that commit crashed
in subprocess._readerthread during the update check (#52649).
Guard every text=True capture site in the desktop-backend process with
encoding='utf-8', errors='replace':
- hermes_cli/web_server.py: git log update panel, memory-provider setup
runner, WhatsApp bridge npm install, docker probe
- hermes_cli/banner.py: all 5 git sites (update check runs at startup)
- tui_gateway/host_supervisor.py: build-sha probe, ps probe, compute-host
Popen drain threads
- tui_gateway/compute_host.py: build-sha probe, ps rss probe
The gateway startup maintenance block gained a maybe_auto_archive call in
the same provably-off-loop __init__ site as maybe_auto_prune_and_vacuum;
bump the reviewed sync-escape count from 3 to 4.
Sessions settings gain an "Auto-archive stale chats" toggle with a
configurable idle threshold, persisted to sessions.* in config.yaml so
the backend sweep owns the policy. Sidebar pins (localStorage) are
mirrored to the backend pinned flag at boot and on every change —
pre-existing pins migrate transparently — so the sweep can never hide a
pinned chat.
New sessions.auto_archive / auto_archive_days config: soft-hide (never
delete) sessions with no activity for N days, aging on last activity
rather than creation so an old-but-active chat is spared. Sweeps are
throttled through state_meta and fire from CLI startup, gateway startup
+ hourly housekeeping, and the serve/dashboard backend (opportunistic
on session list + an hourly lifespan ticker), so every surface honours
one setting.
A new pinned column (declaratively migrated) exempts sessions from the
sweep; PATCH /api/sessions/{id} accepts pinned and flips the whole
compression lineage as a unit, mirroring set_session_archived.
Group the flat recents list and entered-project lanes by recency: an
unlabelled head of the newest run of sessions (cut at a real break in
activity, sized toward the most recent handful), then one divider per
coarse calendar range — Earlier today / Yesterday / Earlier this week /
Last week / Earlier this month / month / month + year. Empty ranges are
skipped, the first rendered group is never labelled, branch clusters
never split, and hand-ordered lists / pinned / project previews stay
divider-free.
Nine bundled and optional skills had stale flags, install URLs, packages, and paths. Verified each against upstream and corrected:
- vllm: removed bogus --enable-metrics/--metrics-port (metrics at /metrics on API port); --speculative-model -> --speculative-config; canonical HF model IDs
- lm-evaluation-harness: --tasks list -> lm-eval ls tasks; --allow_code_execution -> --confirm_run_unsafe_code
- weights-and-biases: wandb.keras import removed -> wandb.integration.keras (WandbMetricsLogger); log_uniform -> log_uniform_values for raw values
- huggingface-hub: upload-large-folder now deprecated; hf papers list -> ls
- openhue: Linux install 404 -> openhue_Linux_x86_64.tar.gz tarball (release repo openhue/openhue-cli, v0.24)
- apple-notes: memo notes -a is a bare flag, no positional title
- excalidraw: upload.py path skills/diagramming/... -> skills/creative/...
- searxng-search: removed Method 3 (searxng-data pip package is a PyPI 404)
- sketch: noted get-shit-done upstream is archived/unmaintained
The design-md skill documented the Apr 2026 (0.1.x) CLI behavior, which
has since drifted:
- Lint rules: the skill listed 7 rules that no longer exist by those
names (duplicate-section, invalid-color, wcag-contrast,
unknown-component-property); the 0.3.0 linter runs 9 rules
(contrast-ratio, orphaned-tokens, missing-primary, missing-typography,
section-order, unknown-key, token-summary, missing-sections,
broken-ref). Verified against live lint output.
- Colors: any CSS color is now valid (oklch/rgb/named), not hex-only.
- Export: json-tailwind (v3) + css-tailwind (Tailwind v4 @theme CSS)
formats; 'tailwind' is a back-compat alias. New exit-code semantics
(export exits 0 regardless of source lint findings).
- Section order / duplicate headings are lint warnings, not file
rejection (verified: duplicate + out-of-order sections exit 0).
- Windows: documented the designmd dot-free bin alias (the design.md
bin name collides with the .md file association); skill declares
platforms: [windows].
- New pitfall: typography sub-property typos (fontwight) are silently
dropped with no finding as of 0.3.0.
All claims verified by running @google/design.md 0.3.0 live (lint,
export, duplicate-section, oklch token, starter template lints clean).
Docs page regenerated via generate-skill-docs.py.
Four coding-agent CLI skills drifted from their live CLIs. Verified against live --help/npm and corrected:
- codex: --full-auto deprecated -> --sandbox workspace-write; --yolo -> --dangerously-bypass-approvals-and-sandbox (yolo kept as noted alias)
- claude-code: --effort levels low/medium/high/xhigh/max (dropped removed 'auto', added 'xhigh'); fixed stray table cell
- grok: --session-id is UUID-only for new sessions (cannot resume by name); rewrote the Session Continuation example; noted --max-turns now exists
- blackbox: wrong npm package (@blackboxai/cli is unrelated) -> @blackbox_ai/blackbox-cli; removed dead source-repo link and phantom session/info subcommands
session.branch wrote the child ROW into the parent's profile db but
built the live agent with the launch defaults: _make_agent fell back to
_get_db() and no HERMES_HOME override was active. The branched agent's
own message flushes — and any later compression rotation it performed —
therefore landed back on the launch profile, splitting the lineage one
turn after the branch. Mirror session.create/resume: open the parent
profile's SessionDB for the agent and hold the home override across the
build, so config/skills/memory resolve to the profile too.
Spotted in #70605's sibling implementation of the same fix.
Co-authored-by: HexLab98 <liruixinch@outlook.com>
The blanket MAX_DESCRIPTION_LENGTH=1024->60 change is narrowed:
create-time validation now rejects new skills whose description
exceeds SKILL_PROMPT_DESC_LIMIT (60) with actionable guidance, while
edit/patch paths stay permissive (warning via system_prompt_preview)
so existing over-limit skills remain maintainable. Runtime display
truncation in skills_tool is left at 1024 (display behavior is a
separate concern from authoring validation).
Boundary tests: 60 accepted, 61 rejected at create; edit/patch on
over-budget skills still succeed.
MAX_DESCRIPTION_LENGTH was set to 1024, but the documented skill-
authoring standard specifies <=60 characters. The model generates
descriptions up to 202 chars because the validation allows 1024.
Lower MAX_DESCRIPTION_LENGTH from 1024 to 60 to match the documented
standard. The system-prompt skill index already truncates to 60 chars,
so over-length descriptions lose their routing signal past char 60.
Fixes#52367
Widening pass on top of the #38820 salvage: a full-graph audit of every
SKILL.md (bundled + optional) found 13 more references to skills that
no longer exist. Classes:
- deleted in the 38d3c49aaf bundled-skill cleanup: generative-widgets,
spotify, cloudflared-quick-tunnel, webhook-subscriptions,
debugging-hermes-tui-commands -> dropped
- native-mcp absorbed into the hermes-agent hub skill -> re-pointed
- toolset names that were never skills: browser, image_gen -> dropped
Audit now reports zero broken related_skills references.
Salvaged from PR #38820 by @bedirhancode, re-applied at current skill
locations (obliteratus and s6 moved to optional-skills/ since the PR):
- research-paper-writing: drop ml-paper-writing (never existed)
- touchdesigner-mcp: drop native-mcp (consolidated) + hermes-video (never existed)
- obliteratus: vllm -> serving-llms-vllm, gguf -> llama-cpp
- s6-container-supervision: drop hermes-agent-dev (not a repo skill)
4 of the original 8 hunks were dropped: heartmula already fixed in
#70453; native-mcp SKILL.md deleted from main; architecture-diagram and
comfyui hunks removed refs to concept-diagrams and
stable-diffusion-image-generation, which are valid optional skills.
Auto-gen page slugs, catalog rows/paths, sidebar entries, and zh-Hans
mirrors follow the directory renames. Also updates the install path
official/creative/audiocraft -> official/creative/audiocraft-audio-generation
in the songwriting-and-ai-music pointer section.
Salvaged from PR #42788 by @Love-JourneY, re-applied at current locations
(audiocraft and segment-anything have since moved to optional-skills/):
- skills/mlops/inference/vllm -> serving-llms-vllm
- skills/mlops/evaluation/lm-evaluation-harness -> evaluating-llms-harness
- optional-skills/mlops/models/segment-anything -> segment-anything-model
- optional-skills/creative/audiocraft -> audiocraft-audio-generation
Directory name != frontmatter name breaks skill_view() lookup by dir
name and causes hermes update sync re-seeding duplicates (#42786).
The authoring guide calls this out as Pitfall #8.
Fixes#42786
RFC 8628 §3.2 makes the device-authorization `interval` optional with a
client-side default of 5 seconds. request_device_code required it, so a
compliant AS that omitted it hit the malformed-response path and the flow
could never complete. Fall back to 5s and cover it with a regression test.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Adds a device authorization grant flow alongside the existing loopback
OAuth flow, so `hermes setup` can connect to Honcho cloud from SSH and
other no-browser environments.
- oauth.py: new HTTP seams — _http_post_form_status (non-raising, since
RFC 8628 polling reads the OAuth error off a 400) and _http_get_json
for the RFC 8414 metadata probe
- oauth_flow.py: DeviceCode, request_device_code, poll_for_token with
slow_down backoff (+5s, capped at 60s) bounded by expires_in, typed
errors (AccessDenied, DeviceCodeExpired, AuthorizationTimeout), and
supports_device_login (fail-closed metadata gate); device flow ends in
the same install_grant tail as loopback so refresh/status work
unchanged
- oauth_flow.py: loopback callback now serves a "sign-in was not
completed" page on consent cancel instead of the success page
- cli.py: cloud menu offers oauth / device / apikey; the device option
only appears when the host advertises the grant, and becomes the
default when no browser is detected
- 18 new tests covering the full flow against a local fake AS, backoff
schedule, error mapping, deadline bound, metadata gate, and wizard
branches
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The multiplex cron path only used use_cron_store() to scope storage paths
(jobs.json, heartbeat files), but _get_lock_paths() and the agent execution
path in cron/scheduler.py resolve via _get_hermes_home() → get_hermes_home()
which checks _HERMES_HOME_OVERRIDE, a separate ContextVar. Without
set_hermes_home_override(), the .tick.lock, config.yaml, .env, and secrets
all resolved to the default profile instead of the per-profile home.
This matches the web_server.py pattern (line 11994) which sets both
set_hermes_home_override(home) AND use_cron_store(home), and the
_profile_runtime_scope pattern used for the multiplexed inbound path.
Found via 3-agent parallel review of salvaged PR #69529.
Under multiplex_profiles, the gateway starts a single InProcessCronScheduler
bound to the process-global HERMES_HOME (the default profile's home), so
only that profile's cron/jobs.json is ticked. A job registered from a
secondary-profile session lands in <profile>/cron/jobs.json, reports a valid
next_run_at — and never fires.
Changes:
1. cron/scheduler_provider.py — InProcessCronScheduler.start() now accepts
an optional profile_homes kwarg (list of (name, Path) tuples). When set,
_start_multiplex() iterates tick() over each profile home using
use_cron_store(), so every served profile's cron store is ticked on
every tick cycle. Heartbeats and interrupted-execution recovery are also
scoped per profile via use_cron_store().
2. gateway/run.py — start_gateway() now resolves profiles_to_serve(multiplex=True)
when multiplex_profiles is on and passes them to the cron scheduler as
profile_homes. Only applies to InProcessCronScheduler (the built-in);
external providers are unchanged.
3. cron/jobs.py — record_ticker_heartbeat(), get_ticker_heartbeat_age(), and
get_ticker_success_age() now resolve paths via _current_cron_store()
instead of module-level TICKER_HEARTBEAT_FILE / TICKER_SUCCESS_FILE
constants. This makes heartbeats correctly scoped per profile, so
'hermes cron status' reflects liveness for every profile independently
under multiplex_profiles.
4. tests/cron/test_scheduler_provider.py — two new tests:
- test_multiplex_ticker_ticks_each_profile_once: verifies tick() is called
once per profile per tick cycle.
- test_multiplex_heartbeat_scoped_per_profile: verifies heartbeat files
are written to each profile's cron store.
Widens the salvaged .env injection fix (#50315) to the sibling site it
missed: hermes_cli/memory_setup.py::_write_env_vars is the near-identical
core writer the openviking plugin's copy was forked from, is fed directly
by interactive _prompt() (pasted API keys), and is reused by other memory
plugins (e.g. supermemory imports it). A pasted secret with an embedded
CR/LF injected an arbitrary extra KEY=VALUE line on the next read.
Same _env_line_safe() treatment as the plugin writer (strip every
str.splitlines() separator + NUL), matching config.save_env_value's
existing newline strip. Mutation-checked: reverting the sanitizer makes
the new regression tests fail.
Follow-up hardening on the salvaged _ensure_client() (#21130 fix):
- Failed-config cooldown: after a refresh attempt fails for a given
resolved config, skip re-probing for 30s. Previously every provider
access against a down endpoint paid a 3s health probe under
_client_refresh_lock and emitted a warning (2+ per turn, some on
user-facing threads: prefetch, tool calls, session end). Retries
still happen after the cooldown or immediately when config changes,
and the log message now says so instead of the false 'disabled until
config changes'.
- Atomic connection snapshot: _conn_snapshot (5-tuple, single
assignment) is published only after a health check passes.
_new_client() and on_memory_write's writer read it as one load, so
background writers can no longer observe a torn mix of old/new
identity fields mid-refresh or target an endpoint that never passed
health. Field writes in _ensure_client_locked keep tracking the
attempted config for the unchanged-config dedupe.
- _env_refresh_enabled moves to the top of initialize(): an exception
mid-initialize (swallowed by MemoryManager) can no longer leave the
provider silently stuck in never-refresh mode.
- _search_prefetch_context reuses _new_client() and degrades to ''
on construction failure instead of propagating.
Mutation-checked: neutering the cooldown or publishing the snapshot on
failed health makes the new regression tests fail.
Avoid spawning multiple local OpenViking server processes while a runtime autostart waiter is already active. Remote endpoints still retry on later accesses because they do not install a local waiter.
Route refreshed unreachable local OpenViking configs through the existing runtime recovery path so /reload can attach to a locally starting server instead of disabling memory until restart.
(cherry picked from commit 040e18ad907a55442c54ea59db16d4ca3f61daa9)
The _needs_trusted_identity_retry method was hard-coding specific
server-side error strings to detect when a request failed due to
missing X-OpenViking-Account / X-OpenViking-User headers. Each new
server-side error variant required another string added to the client.
Replace the string enumeration with a structural match: the error
message mentions one of the tenant headers AND the HTTP status is 400.
This covers all current error variants:
- "Trusted mode requests must include X-OpenViking-Account and User"
- "ROOT requests to tenant-scoped APIs must include X-OpenViking-Account"
- "Trusted mode requests must include X-OpenViking-Account."
- "Trusted mode requests must include X-OpenViking-User."
The 400 status guard avoids false-positives on 403 errors such as
"USER API keys cannot override X-OpenViking-User", which must not
trigger a retry.
All 176 existing tests pass.
(cherry picked from commit 5a24d6766ce1ae89b20266b751c5746a8243c378)
`OpenVikingMemoryProvider.shutdown()` joins in-flight writers, deferred-commit
threads, and prefetch threads, but not `_runtime_start_thread` — the tracked
`daemon=True` waiter that runs `_finish_runtime_openviking_start`, which blocks
on network health probes (`_wait_for_openviking_health` polling + a
`_VikingClient.health()` request).
If the local OpenViking runtime is slow or unreachable, that waiter can still
be blocked in network I/O at interpreter exit. CPython then forcibly kills it
during `Py_FinalizeEx` (`PyThread_exit_thread` -> `__pthread_unwind` ->
`abort()`), producing SIGABRT (exit 134) with no traceback — the same daemon-
thread-at-exit failure class fixed for the Honcho provider.
Fix:
- `shutdown()` now joins `_runtime_start_thread` (timeout-bounded) alongside the
other tracked threads.
- `_wait_for_openviking_health()` gains a `should_stop` callback; the waiter
passes `lambda: self._shutting_down` so the poll loop bails out promptly once
`shutdown()` flips the flag, instead of lingering up to the 60s autostart
timeout and timing out the join (which would leave the thread alive).
- Add tests/plugins/memory/test_openviking_shutdown.py covering the short-circuit
and the shutdown-joins-runtime-thread behaviour.
(cherry picked from commit 5471ec70210b35462450aa52f0cd483439fffba7)
initialize() snapshots OPENVIKING_* into the provider once, so /reload
(which only updates os.environ) leaves viking_* tools running against
stale auth — users have to restart hermes to pick up keys added to
~/.hermes/.env after startup.
Add _ensure_client(), which re-resolves the connection settings via the
same _resolve_connection_settings/_load_hermes_openviking_config path
initialize() uses and rebuilds + health-checks the client only when an
OPENVIKING_* value actually changed; otherwise it reuses the cached
client so the hot path stays at one dict comparison with no network
calls. Every `if not self._client:` guard in system_prompt_block,
queue_prefetch, sync_turn, on_session_end, on_memory_write and
handle_tool_call now goes through it.
Refreshing is gated behind a flag set at the end of initialize() so the
baseline is established before any env re-resolution happens — callers
that wire up a client directly keep the existing client untouched.
Refs #21130
(cherry picked from commit b694d21b7c4ff0330df6051e12dc8991f7ea10a6)
`_write_env_vars` in the OpenViking memory provider interpolates each
secret straight into a `KEY=VALUE` line, but the values only ever pass
through `_clean_config_value`, whose `value.strip()` trims surrounding
whitespace and leaves internal CR/LF intact. Because the file is strictly
line-oriented and is re-read via `read_text().splitlines()`, a value that
carries an embedded newline spills onto a second physical line, and the
tail is re-parsed as an independent `KEY=VALUE` entry on the next round
trip. A secret pasted with a trailing record (e.g. an `OPENVIKING_API_KEY`
copied with an extra line) therefore injects an arbitrary additional
variable into the persisted credentials file and silently corrupts it.
The fix neutralizes the line terminators at the single chokepoint where
values reach the file. A small `_env_line_safe` helper strips `\r`, `\n`,
and the NUL byte from each value, and both write sites in `_write_env_vars`
(the existing-key update branch and the appended-key branch) route through
it, so a value can only ever occupy the single line it is written on.
## What does this PR do?
Hardens the OpenViking memory provider's `.env` writer so a malformed or
pasted secret value can no longer break out of its `KEY=VALUE` line and
inject a rogue variable into the profile-scoped credentials file.
## Related Issue
N/A
## Type of Change
- [x] 🐛 Bug fix (non-breaking change that fixes an issue)
## Changes Made
- `plugins/memory/openviking/__init__.py`: add `_env_line_safe()` which
removes `\r`, `\n`, and `\x00` from a value, and apply it to both the
updated-key and appended-key write branches in `_write_env_vars()`.
- `tests/plugins/memory/test_openviking_provider.py`: add two regression
tests covering a fresh write and an in-place key update with embedded
CR/LF, asserting no injected line survives the read-back.
## How to Test
1. Run the targeted tests:
`pytest tests/plugins/memory/test_openviking_provider.py -k env_writer -q`
2. Reverting the `_env_line_safe` sanitization makes
`test_openviking_env_writer_strips_embedded_newlines_in_values` and
`test_openviking_env_writer_strips_newlines_when_updating_existing_key`
fail with a rogue `INJECTED_KEY=`/`ROGUE=1` line appearing in the file,
confirming the tests pin the bug.
3. `ruff check plugins/memory/openviking/__init__.py` and
`python scripts/check-windows-footguns.py plugins/memory/openviking/__init__.py`
both pass.
## Checklist
### Code
- [x] I've read the Contributing Guide
- [x] My commit messages follow Conventional Commits
- [x] I searched for existing PRs to make sure this isn't a duplicate
- [x] My PR contains only changes related to this fix
- [x] I've run the relevant tests and they pass
- [x] I've added tests for my changes (required for bug fixes)
- [x] I've tested on my platform: macOS 15 (Darwin 25.5)
### Documentation & Housekeeping
- [x] I've updated relevant documentation (docstrings) — or N/A
- [x] I've updated `cli-config.yaml.example` if I added/changed config keys — N/A
- [x] I've updated `CONTRIBUTING.md` or `AGENTS.md` if I changed architecture or workflows — N/A
- [x] I've considered cross-platform impact (strips CR as well as LF) — done
- [x] I've updated tool descriptions/schemas if I changed tool behavior — N/A
(cherry picked from commit f29dd2df845c3f049820797b473437b577b75362)
profile_name was only written on the agent's initial lazy create
(e8b7ce8c1); every parented child row — compression rotation, TUI
/branch, desktop branch first-persist — was created without it. A
non-default profile's lineage therefore turned NULL on its first
compression or branch and aggregated as "default" in unified session
lists, completing the cross-profile session-jump.
Fix the class at the DB layer: _insert_session_row's parent backfill now
COALESCEs profile_name from the parent alongside cwd/git_* (#64709
pattern), so any parented child inherits its lineage's owning profile.
Stamp it explicitly at the three create sites as well — compression
rotation (mirroring _ensure_db_session), TUI session.branch, and the
TUI first-prompt row persist — so rows are self-describing even when the
parent row predates the profile_name column.
forkBranch called session.create without a profile, so in app-global
remote mode (one backend serving every profile) a branch of a
non-default-profile chat silently landed on the launch profile — the
"session jumps between profiles after branching" bug. New chats already
carry their profile (desktopSessionCreateParams, #39993); branching was
the remaining create callsite without it.
Thread the parent's owning profile through both branch entry points:
branchStoredSession already resolved it (#67603) but dropped it before
the create; branchCurrentSession now resolves the open chat's profile
via the same cache → active → cross-profile ladder as resume. forkBranch
swaps the live gateway onto that profile and passes it on session.create,
which also makes the optimistic sidebar row's profile stamp correct.
Supersedes #40788 — same invariant, rebuilt for the current forkBranch/
branchStoredSession split.
Co-authored-by: Dusk1e <yusufalweshdemir@gmail.com>
Closes#62503
Root cause: session.resume already opened the requested profile's state.db,
but session.list / most_recent / delete / history / title / status, create's
immediate info payload, teardown, session.branch, and post-turn pending_title
still used the launch profile (_get_db / _current_profile_name).
Fix: add _db_for_profile / _profile_db / _session_db helpers and route
session.* methods through them; report requested profile on create/status;
lifecycle uses session-owned profile stores with branch profile_home inherit.
Rebase onto current main keeps _session_usage_snapshot in session info.
Audited every reference against ground truth (COMMAND_REGISTRY, argparse
--help output, TOOLSETS dict, provider profiles, DEFAULT_CONFIG) and
corrected drift the restructure had carried over verbatim:
- slash-commands.md: rebuilt from the live registry. Removes the phantom
/skill command (never existed — skills load via /skills, hermes -s, or a
skill's own /<name> command) and the wrong /q alias on /quit (/q belongs
to /queue) — both reported by @liuhao1024 in #50608 and @gauravsaxena1997
in #50613. Adds ~25 real commands that were missing (/learn, /memory,
/pet, /hatch, /bundles, /moa, /suggestions, /blueprint, /whoami, …) with
correct aliases and CLI/GW scoping.
- cli-reference.md: verified flags (-z/--oneshot, --tui/--cli, --safe-mode),
real subcommand sets for config/setup/sessions/skills/gateway/mcp/profile/
auth/webhook/cron/skin/pets, added missing top-level commands (fallback,
project, moa, logs, console, hooks, security, backup); condensed tables.
- providers-and-models.md: rebuilt from the 36 shipped provider profiles
with their actual env vars (was 21 rows, several stale); folds in the
alias system and fallback-chain command.
- configuration.md: toolset table matches TOOLSETS (adds coding,
computer_use, video_gen, context_engine, project; drops nonexistent
messaging/rl), config sections match DEFAULT_CONFIG, STT/TTS provider
lists match the shipped set.
- background-systems.md: curator verbs match the argparse surface.
- troubleshooting/security-privacy/windows-quirks/contributor-guide: fixed
dangling 'section above/below' references from the body split; moved the
reset-permissions playbook to security-privacy.md (its routing home).
Reported-by: liuhao1024 <sunsky.lau@gmail.com>
Reported-by: gauravsaxena1997
Flips the default fan-out cadence from per_iteration (advisors re-run on
every tool iteration, multiplying advisor spend by tool-loop depth) to
user_turn (advisors run once on the first message of each user turn; the
acting aggregator works the rest of the tool loop with that turn's
advice). Until per-mode benchmarks justify a costlier default, MoA
defaults to the cheapest, lowest-impact cadence (#67199).
One default for everyone — no split legacy/new-preset semantics; presets
that want per-step advising set fanout: per_iteration explicitly. All
three modes (user_turn / per_iteration / every_n:N) remain selectable;
every_n:1 still collapses to per_iteration (semantic identity), while
unparseable values now fall to user_turn (the default).
Docs updated with a default-change note; the per-iteration rerun test
pins its mode explicitly.
Co-authored-by: skyer-flyyy <188930297+skyer-flyyy@users.noreply.github.com>
Every SKILL.md description over 60 chars was silently truncated to
57 chars + '...' in the system-prompt skill index
(extract_skill_description, agent/skill_utils.py), destroying the
routing signal for 69 of 179 skills — some descriptions ran to
1,005 chars.
Rewrites follow the authoring standard: <=60 chars, one sentence,
ends with a period, trigger front-loaded, no marketing words, no
skill-name repetition. Excess detail already lives in each skill's
body.
Includes the touchdesigner-mcp trim from PR #32361 (credit:
@JeliTron) and aligns with the router-precision direction of PR
#48780 (@John-Lussier). Docs catalogs + per-skill pages regenerated
via website/scripts/generate-skill-docs.py.
Co-authored-by: JeliTron <287797501+JeliTron@users.noreply.github.com>
The skill-authoring guide and curator prompt both reference
descriptions as the primary discovery mechanism but never mentioned
the 57-char system prompt truncation. Add explicit guidance:
- Authoring guide: frontmatter docs, template comment, size limits,
pitfall #3 with good/bad examples, verification checklist
- Curator prompt: parenthetical noting the 57-char window when
writing umbrella skill descriptions
When a skill is created or edited with a description longer than
SKILL_PROMPT_DESC_LIMIT (60 chars), the tool response now includes a
system_prompt_preview field showing exactly what the system prompt
skill index will display. This gives the agent immediate feedback to
self-correct truncated trigger phrases.
Also adds tool schema guidance about the 57-char window and fixes a
stale docstring in skill_commands.py that incorrectly claimed the
system prompt renders the full description.
The system prompt skill index truncates long descriptions to 57 chars,
but this limit was a hardcoded magic number. Extract it as a named
constant and factor the normalization logic into a shared private
helper so the extraction function and the new truncation predicate
cannot drift.
No behaviour change — pure refactor.
Follow-up on the salvaged commit: the xurl skill loads even when x_search
isn't registered (check_fn-gated on xAI credentials), so per the
cross-toolset reference rule the skill must not name it. Replaced the
skill's x_search routing block and workflow step with skill-native
wording (raw engageable posts, authenticated context, write-evidence
rule). The cross-surface comparison stays on surfaces where both are
known to exist: x-search feature docs, toolset description, tools-config
setup note, and the x_search tool schema (kept generic, no tool names).
Rewrote the routing tests to pin the placement contract, including that
the skill never names credential-gated surfaces.
Make the x_search / xurl boundary explicit in the skill, feature docs,
toolset metadata, setup note, and reference pages, while keeping the
model-facing x_search schema generic (no static xurl name).
Regression tests assert behavioral routing invariants rather than frozen
prose snapshots. Drop the stale CI-only plugin/hangup hunks already on
main so this rebases cleanly.
skills/computer-use/SKILL.md sat at the root of the bundled skills
tree as an uncategorized single-skill directory. Move it to
skills/autonomous-ai-agents/computer-use/ alongside claude-code,
codex, opencode, and hermes-agent.
- git mv skills/computer-use -> skills/autonomous-ai-agents/computer-use
(history preserved)
- root-level-skill docstring examples (commands.py,
generate-skill-docs.py) now use a generic placeholder instead of
naming a specific skill, since categorizing root-level skills is
ongoing
- docs: move the bundled page to autonomous-ai-agents-computer-use,
update sidebars.ts, skills-catalog.md, and the two skill-path
references in features/computer-use.md
E2E validated: fresh sync_skills() copies to the new nested path;
existing installs with the old flat copy keep it untouched (manifest
name-keyed, hash unchanged -> skipped, no duplicate);
_get_category_from_path resolves 'autonomous-ai-agents'; docusaurus
build green; 396 targeted tests pass.
The 'queues an Enter-submitted draft while compaction is active' test
pastes a large message to push the session over the fixture's 22k
threshold_tokens. At repeat(500) the payload is only ~4k tokens — the
other ~18k came from the ambient system prompt (tool schemas + skills
index + memory), leaving the trigger margin-less. The hermes-agent
skill hub restructure (e3d524b482) shrank the bundled skills index by
~160 tokens and dropped the total just under threshold: compaction never
started, waitForHeldCompletion() hung, and the test timed out at 90s on
every branch since — including main (first red run: 9a4d1a0130).
Bump the payload to repeat(1500) (~12.4k tokens, total ~30k) so the
test crosses the threshold on its own weight with ~8k tokens of margin,
and document the invariant so the next prompt-weight change doesn't
resurrect this.
A direct run (cronjob action='run' / webhook-triggered manual fire) takes
the job's fire_claim and advances next_run_at. When it races the external
provider's scheduled fire for the same occurrence, Chronos loses the claim
and — by design — does not re-arm (the winner owns the re-arm). But the
direct-run winner never notified the provider, so the NAS one-shot for the
consumed occurrence was left stale forever and the recurring job silently
stopped firing.
Observed in production: a managed 1-minute review job stalled for 20 hours
because a GitHub-webhook direct run claimed the job 2s before the Chronos
fire arrived; every subsequent occurrence was orphaned while /api/status
stayed green.
Fix: after a *claimed* direct execution completes (success or failure —
next_run_at advances at claim time either way), call
_notify_provider_jobs_changed_safe() so the active provider re-arms the
post-run next_run_at. No-op for the built-in ticker; claim-lost direct
runs still never notify (the winning scheduler owns the re-arm).
Claude Code's /simplify was renamed away, community-restored, and rebuilt
since our 3-agent port (v2.1.63 -> v2.1.154+). This brings simplify-code
in line with the current upstream design, re-expressed in our own wording
and layered onto our existing risk-tier/confidence machinery:
- New Reviewer 4 (Altitude): flags band-aid fixes layered on shared
infrastructure — special cases, symptom patches with unfixed sibling
sites, workaround stacks — and points at the deeper mechanism fix.
Matches our AGENTS.md 'fix the class, not the site' rubric.
- Explicit cleanup-vs-bug-hunt boundary: this skill improves working
code; correctness review stays with requesting-code-review.
- Inline single-pass fallback when delegate_task is unavailable (leaf
subagents, delegation disabled) — previously the skill just broke;
now all four angles run sequentially with honest disclosure.
- Finding format gains a concrete-cost field.
- Efficiency reviewer adds closure-capture scope-retention leaks.
- Pitfalls: fan-out cap 3 -> 4, band-aid-vs-deliberate-boundary caveat,
no-bug-hunting drift guard.
Kept our value-adds upstream lacks: SAFE/CAREFUL/RISKY tiers,
Chesterton's Fence via git blame, dry-run/focus/scope modifiers.
With the built-in x_search tool (xAI-backed, returns a synthesized
answer) now shipping, the xurl skill's plain 'searching posts' wording
no longer disambiguated the two search surfaces. Describe xurl search
in its own terms — authenticated X index query returning raw post JSON
with IDs suitable for immediate engagement — so agents route correctly
whether or not x_search is enabled. No cross-toolset reference added,
per the schema/skill cross-reference rule.
Add two pitfalls discovered when running the skill against a fresh
Jupyter server:
- Pitfall #9: When the websocket reply channel hangs on every execute
even though the kernel actually ran (REST shows execution_state=idle
and execution_count increments), force zmq transport with
--transport zmq. The zmq transport uses jupyter_client directly and
sidesteps the broken websocket layer.
- Pitfall #10: A fresh ServerApp rejects POST /api/sessions with
"_xsrf argument missing from POST" unless you start it with
--ServerApp.disable_check_xsrf=True. Needed for REST-only flows
where no browser/cookie is establishing the XSRF token.
The yuanbao skill is platform-specific guidance for Tencent Yuanbao
group chats — niche for the default bundled set. Per the
'when in doubt, optional' rule, ship it as an optional skill.
Install via: hermes skills install official/yuanbao/yuanbao
The nano-pdf description repeated the skill name ('via nano-pdf CLI'),
violating the skill authoring standard, and spent its char budget
restating the name instead of distinguishing the skill from the
structural pdf skill. New description: 'Edit text in existing PDFs via
natural-language prompts.' (56 chars) — no name repetition, and
'text in existing PDFs' contrasts cleanly with pdf's
'Create, merge, split, fill, and secure PDF files.'
Propagated to the auto-generated docs pages and zh-Hans locale
(catalog row + per-skill page) for locale parity.
Per the 'when in doubt, optional' rule — zero-shot image segmentation
via SAM is a niche computer-vision capability with heavy deps
(torch, transformers), not something most users load.
Install via: hermes skills install official/mlops/segment-anything
Both are heavy-dep, GPU-bound local music-generation skills (8-16GB VRAM,
torch stacks, manual source patches) that were bundled in two different
categories (media/ and mlops/models/). Per the 'when in doubt, optional'
rule they now sit side by side in optional-skills/creative/, and the
bundled songwriting-and-ai-music skill points at them for local generation.
- git mv skills/media/heartmula -> optional-skills/creative/heartmula
- git mv skills/mlops/models/audiocraft -> optional-skills/creative/audiocraft
- cross-link related_skills both ways + songwriting-and-ai-music
- new section 10 in songwriting-and-ai-music with install commands
- docs: bundled pages -> optional pages (en + zh-Hans), catalogs, sidebar
Install via:
hermes skills install official/creative/heartmula
hermes skills install official/creative/audiocraft
The hermes-agent skill body was a 51KB monolith loaded in full on every
trigger. It is now a lightweight hub (~12KB): identity, quick start, the
tmux orchestration guide (kept in-body — the autonomous-ai-agents category
contract), surface orientation, and hard invariants, with a routing table
into 18 reference files that carry the depth.
Absorbed four Hermes-specific skills as first-class references so their
content gains a discoverable home and room to grow without bloating any
primary body:
- skills/hermes-themes -> references/themes.md + templates/skin.yaml
- skills/hermes-desktop-plugins -> references/desktop-plugins.md + templates/plugin.js
- skills/productivity/tui-widgets -> references/tui-widgets.md + templates/clock.mjs
- skills/productivity/petdex -> references/petdex.md
New references extracted from the old body: cli-reference, slash-commands,
providers-and-models, configuration, project-context-files,
security-privacy, background-systems, windows-quirks, troubleshooting,
contributor-guide (native-mcp and webhooks already existed). Also commits
delegate-task-concurrency-diagnosis and portal-auth-for-third-party-apps,
which the old body referenced but the repo never shipped.
Description updated to cover the widened scope:
'Use, configure, theme, extend, and orchestrate Hermes Agent.' (60 chars)
skills/dogfood/SKILL.md sat at the root of the bundled skills tree,
making it one of the few uncategorized skills (Discord /skill
autocomplete listed it under 'uncategorized'; hermes_cli/commands.py
cited it as the example). Move it to
skills/software-development/dogfood/ alongside the other QA/testing
skills (test-driven-development, systematic-debugging,
requesting-code-review).
- git mv skills/dogfood -> skills/software-development/dogfood
(history preserved)
- fix test fixture path in tests/tools/test_browser_console.py
- update root-level-skill docstring examples (commands.py,
generate-skill-docs.py) to cite computer-use, which is still root-level
- drop 'dogfood' from the category list in hermes-agent-skill-authoring
SKILL.md (en + zh-Hans docs mirrors)
- docs: regenerate/move the bundled page to
software-development-dogfood (en + zh-Hans), update sidebars.ts,
skills-catalog.md, and the adversarial-ux-test related-skills link
E2E validated: fresh sync_skills() copies to the new nested path;
existing installs with the old flat copy keep it untouched (manifest is
name-keyed, hash unchanged -> skipped, no duplicate);
_get_category_from_path resolves 'software-development'; docusaurus
build green.
Maintainer scoping decision for the #51226 salvage: document that
select_context() is for engines that must REPLACE per-request context
(retrieval/routing) — pre_llm_call is inject-only by documented cache
design; that observation-only plugins should implement a MemoryProvider
(sync_turn) rather than a context engine, with on_turn_complete scoped
as the observation mirror for engines that already select; and that a
non-no-op select_context naturally changes the prompt-cache prefix on
turns where the selection changes — engines should return stable
selections when nothing changed.
Verification follow-up for the #51226 salvage: the host call site guarded
select_context with hasattr(), but the ABC defines a default on every
engine, so the built-in ContextCompressor (and any non-implementing
engine) still paid per-request shallow copies of the conversation
history plus a hook call on every provider request. Identity-check the
bound method against ContextEngine.select_context and return the
request untouched — mirroring the existing base-method short-circuit in
_notify_context_engine_turn_complete — so the default path does zero
work, not just produces an identical result.
Adds two pins: the base no-op is never invoked (patched-to-raise base
stays silent), and ContextCompressor.__dict__ contains neither new verb.
Also registers the contributor email mapping for @chaos-xxl.
- website guide: on_turn_complete() now carries the same best-effort coverage
caveat as the ABC docstring (fires from the finalization seam; abnormal
early-return paths bypass it) — removes the doc/code inconsistency.
- test: finalization seam emits on_turn_complete with usage=None + the
interrupted flag for an interrupted finalized turn. Docstring records that
the negative early-return-bypass half is best-effort and deferred to a
shared-seam follow-up rather than pinned via a full run_conversation harness.
Addresses the hermes-sweeper review on #51226:
- _apply_context_engine_selection now passes shallow copies of the read-only
conversation_messages / incoming_message to the hook, so an engine mutating
them in place cannot corrupt persisted transcript state (enforces the
request-only contract, not just documents it). Adds a mutation-regression
test asserting persisted history + incoming message are untouched.
- on_turn_complete docstring: scope the coverage claim to the standard
finalization seam. Some abnormal early-return paths (content-policy block,
provider terminal failure) currently persist+return without finalization and
don't emit the hook; documented as best-effort with a shared-seam follow-up,
rather than over-promising a guaranteed callback for every early exit.
- _apply_context_engine_selection: reject an empty list. all([]) is True, so
a [] returned by a failing/buggy engine previously replaced a valid request
with an empty message list the downstream sanitizers can't restore; now it
falls open to the unmodified request (honors the fail-open contract).
Thanks @johnnykor82 for catching this on #41918's review.
- test: empty list keeps the original request (fail-open regression).
- docs: document select_context()/on_turn_complete() in the public
context-engine plugin guide (were still describing only the old contract).
- context_engine.py: document that select_context() runs before cache-control
and all request sanitizers, so (a) replacements still pass host validation
and (b) the no-op default keeps the request byte-stable (AGENTS.md prompt-
cache invariant). Note the hook is evaluated per provider request.
- tests: no-op path is byte-stable for cache-control; a role-unusual
replacement is passed through for the existing downstream sanitizers to
normalize (select_context does structural validation only).
The on_turn_complete() observation hook is the engine's post-turn signal,
so it should receive the completed turn's canonical token usage when the
host has it, not a hardcoded None. Per @johnnykor82's #41918 contract: the
engine uses prompt/completion + cache_read/write/reasoning buckets to judge
how large/expensive the selected context was before the next select_context().
- conversation_loop.py: stash the most recent provider response's usage_dict
(the same canonical shape fed to update_from_response) on the agent as
_last_turn_usage; reset to None at turn start so turns that never reach a
provider response (early failure / interrupt) forward None, not a stale
prior turn's usage.
- turn_finalizer.py: forward agent._last_turn_usage instead of usage=None.
- context_engine.py: document the usage param contract on the ABC hook.
- tests: cover both ends through the real finalize_turn path — completed turn
forwards the full canonical bucket set intact; no-response turn forwards None.
Co-authored-by: johnnykor82 <johnnykor82@users.noreply.github.com>
Adds the post-turn observation verb as the companion to select_context():
an optional, no-op-default on_turn_complete() called once after the
assistant/tool loop finishes, with the finalized transcript snapshot. Lets
an engine ingest/index/summarize the completed turn to inform the next
select_context(). Wired via _notify_context_engine_turn_complete() from
turn_finalizer.finalize_turn(); fail-open, base no-op short-circuited so
non-implementing engines (incl. the built-in compressor) pay nothing.
This is the request-assembly + observation pair from #41918; with this
commit the PR fully subsumes #41918's two hooks (prepare_request_messages
-> select_context, on_turn_complete) rather than only the selection half.
Co-authored-by: johnnykor82 <johnnykor82@users.noreply.github.com>
Adds an optional, no-op-default select_context() hook to the ContextEngine
ABC, called every turn after the request messages are assembled and before
provider dispatch — independent of should_compress(). Lets an engine select
or replace which context enters the prompt for a single request (retrieval,
topic routing, role/branch switching) without mutating persisted history,
removing the need to abuse should_compress()=True as a per-turn callback.
The host call site (_apply_context_engine_selection) is fail-open: a missing
hook, an exception, or an invalid return value leaves the assembled request
untouched. Additive and non-breaking: the built-in compressor and every
existing engine are unaffected.
Consolidates the per-turn request-assembly surface proposed across #41918,
Related: #36765#41918#24949#47109#50053#23837#25115#29370
Routine automatic compression stays silent-by-design on chat platforms
(default unchanged, byte-identical). New opt-in config key
compression.progress_notices (bool, default false) opens a gate on the
gateway noise filter (_prepare_gateway_status_message) that lets ROUTINE
compression progress statuses through to chat surfaces.
- Membership is derived from the #69550 status template constants in
agent/conversation_compression.py (compiled to a literal-escaped regex,
never re-inlined wording), so unrelated noisy statuses (aux failures,
provider retry/rate-limit chatter) stay suppressed even when enabled.
- The compaction completion notice (COMPACTION_DONE_STATUS, #69546
lifecycle 'compacted' edge) already flows through the status path and
passes the filter — no new emit site needed.
- Config wired everywhere: hermes_cli/config.py DEFAULT_CONFIG,
cli-config.yaml.example, gateway raw-YAML read (live, mtime-cached),
gateway hot-reload cache-busting key list, website configuration docs.
- Failure notices and manual /compress feedback remain always-visible;
VISIBLE_COMPRESSION_MESSAGES and emit sites untouched.
Design by @havok-training (issue #52995).
_render_tool_calls only handled dict-shaped entries; a SimpleNamespace-
shaped tool_call (SDK-style stream-stitched responses) rendered as
'[called tool: tool]', silently losing the function name and arguments
from the advisory view. Handle both shapes (including a namespace-shaped
nested function inside a dict entry).
One-hunk hardening salvaged from closed#59712.
Co-authored-by: SquabbyZ <601709253@qq.com>
Reference models may have a smaller context window than the aggregator
(e.g. kimi-k2.7-code @ 262K advising a glm-5.2 @ 1M conversation).
Without context-length protection, a reference whose window is exceeded
gets a hard HTTP 400 from the provider, which _run_reference's
try/except silently converts to a [failed: …] note — the MoA turn
silently degrades to fewer references (#60345).
Redesigned implementation of #60387:
- Estimate AFTER the advisory system prompt is prepended, so the
request that is actually sent is what gets budgeted.
- Reserve output headroom: the preset's reference_max_tokens when set,
else an 8192-token constant, plus a 10% estimator-error fraction.
- Trim on advisory-view boundaries (text-only user/assistant turns; no
tool-result frames to orphan), preserving the system prompt, the
user-first invariant after every pop (never assistant-first), and the
trailing synthetic user turn.
- Cache get_model_context_length per (provider, model) in a per-fan-out
dict shared across the worker threads, so a turn resolves each
window once instead of probing metadata sources
per-reference-per-iteration (failures are cached too).
Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com>
Follow-ups for salvaged #56344:
- A reference that completes between the interrupt check and the reap
keeps its REAL output and accounting (the provider call billed) instead
of being zeroed with a placeholder.
- A reference still in flight at interrupt time gets a placeholder in the
results, but its future now carries a done-callback that folds the
eventual real usage/cost into the facade's pending accounting
(late_accounting_sink -> _record_late_reference_accounting), so billed
spend is never silently dropped. Pending totals are folded (not
overwritten) and guarded by a lock since done-callbacks fire on
executor worker threads.
- Interrupted placeholder results are no longer written into the facade's
turn-scoped reference cache: a cache HIT never re-runs references, so
caching a partial snapshot would replay '[skipped: interrupted by
user]' notes for the rest of the turn. The cache is left empty and the
next create() re-runs the fan-out.
agent/tool_executor.py's concurrent tool batch checks agent._interrupt_requested
and aborts the wait early; agent/moa_loop.py's _run_references_parallel had
no equivalent, so a MoA-enabled turn blocked on ThreadPoolExecutor.result()
until every reference model finished or hit its own individual
auxiliary.moa_reference timeout -- there was no way for the user to abort a
live turn mid-fanout.
Thread an optional `agent` parameter through aggregate_moa_context ->
_run_references_parallel (used when MoA references run alongside the main
model) and MoAClient/MoAChatCompletions (used when the MoA preset itself is
the acting model), then poll concurrent.futures.wait() in
_REFERENCE_POLL_INTERVAL_S slices instead of blocking on future.result() per
reference, checking agent._interrupt_requested each cycle.
Deliberately scoped to interrupt/cancel only -- no new or changed timeout
value, so this doesn't overlap open PRs #53784/#53875 (which lower the
per-reference timeout default but don't add interrupt support). `agent` is
optional and defaults to None, so any caller that doesn't pass it keeps
today's uninterruptible blocking behavior unchanged.
Extends the all-references-failed short-circuit (#56975) to the
persistent `provider: moa` facade path: MoAChatCompletions.create()
previously attached 'use the reference responses below' guidance built
entirely from failure sentinels and called the aggregator with it. Now
an all-failed turn attaches either the sanitized unavailability notice
(loud policy) or nothing (silent policy), and the aggregator — which IS
the acting model — simply acts alone. Advisor accounting for the failed
fan-out is still recorded.
Co-authored-by: liuhao1024 <sunsky.lau@gmail.com>
When every MoA reference model returns a failure (HTTP error, timeout,
etc.) or is skipped by the recursion guard, the one-shot aggregator
synthesis call is now skipped entirely. Previously it would try to
synthesise a wall of failure sentinels, which could block for the full
provider timeout (observed ~6 min on SenseNova) before returning a
non-retryable error that left the session hanging.
The early return carries the sanitized unavailability notice (never raw
provider error text, per the failed-reference containment) so the main
agent loop can still act in single-model mode.
Salvaged from #56975, reworked atop the _is_failed_reference helpers.
Follow-ups for salvaged #53784:
- reference_timeout now defaults to None = no per-preset override, so the
reference fan-out inherits auxiliary.moa_reference.timeout (900s default)
via call_llm's own per-task timeout resolution. The PR's 30.0s default
would have cut off long-thinking advisors mid-response, and its 300s max
cap capped legitimate explicit values — both removed. Explicit per-preset
values are still honored as-is.
- _is_failed_reference also treats '[skipped: …]' recursion-guard notes as
internal sentinels, keeping them out of both aggregator prompts.
- Dashboard/desktop TS types updated to number | null; web_server validator
accepts null/empty as 'inherit'.
Keep MoA reference display events off the machine-readable -Q stdout
surface (platform=cli with tool_progress_mode=off) while preserving them
everywhere else. Extracts the relay into module-level helpers so the
policy is testable.
Salvaged from #67334.
Per the 'when in doubt, optional' rule — the live-kernel workflow needs
uv + JupyterLab + a running server + a cloned hamelnb repo, a niche
setup that shouldn't ship active by default.
Renamed jupyter-live-kernel -> jupyter-notebook (skill name, page slugs,
catalogs, sidebar, zh-Hans mirror, darwinian-evolver related_skills).
Install via: hermes skills install official/data-science/jupyter-notebook
Follow-up to the #67690 salvage (@m4r13y). The PR's tools/env_probe.py
hunk was written against the old capture_output=True _run(); #67964/#67999
rewrote _run to temp-file capture on July 20, so that hunk no longer
applied — but the rewritten _run still lacked creationflags and kept
flashing one console per probe (~5 per kanban worker start) from
windowless parents. Re-implement the one-line fix against the current
shape: creationflags=windows_hide_flags() on the temp-file subprocess.run,
preserving the #67964 grandchild-can't-wedge-the-pipe contract.
Also add the tests the PR didn't ship, in
tests/test_windows_subprocess_no_window_flags.py:
- env_probe._run passes CREATE_NO_WINDOW and keeps temp-file (non-PIPE)
stdout/stderr + DEVNULL stdin
- lazy_deps uv install / pip --version probe / pip install fallback /
ensurepip bootstrap all pass CREATE_NO_WINDOW
- suppress_platform_ver_console: POSIX no-op (platform._syscmd_ver
untouched, win32_ver() still returns), and simulated-Windows stubbing
(echo stub installed, idempotent, never raises)
From windowless processes (the pythonw gateway and the kanban workers it
spawns), three spawn paths flash visible console windows on Windows:
1. tools/env_probe.py::_run() ran its interpreter/pip probes
(python3 / python / pip / 'python3 -m pip' / PEP-668 check, ~5 per
worker start) without creationflags — one console flash per probe.
2. tools/lazy_deps.py had four spawn sites with the same defect:
'uv pip install', the 'pip --version' probe, ensurepip, and the
pip install fallback.
Both now pass creationflags=windows_hide_flags() (CREATE_NO_WINDOW on
Windows, 0 on POSIX) — stdio capture still works because the child is
hidden, not detached.
3. CPython 3.11's platform.win32_ver() unconditionally calls
_syscmd_ver(), which runs 'cmd /c ver' via
subprocess.check_output(shell=True) with no window suppression. Any
dependency touching platform.uname()/version()/platform() at import
time flashes one 'cmd' window per windowless process. New helper
_subprocess_compat.suppress_platform_ver_console() (Windows-only,
never raises) stubs platform._syscmd_ver so win32_ver() falls back to
sys.getwindowsversion().platform_version — verified byte-identical
platform.platform() output on CPython 3.11
('Windows-10-10.0.26100-SP0' either way). Called at the top of
hermes_cli/main.py, right after the hermes_bootstrap guard, before
heavyweight imports.
Verified on Windows 11 by polling EnumWindows at ~15 ms and attributing
new visible HWNDs to the suspect process tree (conhost child presence is
NOT evidence of a visible window — it appears even with
CREATE_NO_WINDOW). Tests: tests/tools/test_windows_native_support.py,
test_env_probe.py, test_lazy_deps.py, test_lazy_deps_durable_target.py —
153 passed; the 3 failures are pre-existing on upstream/main in a
Windows environment (POSIX-only assertions and NTFS chmod semantics).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The per-advisor enabled toggle adds enabled=True to normalized slots;
the JSON-string-parse and per-slot max_tokens tests from sibling
clusters asserted exact dicts. Compare against the enabled-augmented
expectation instead.
The per-reference-model enabled toggle (#59753 salvage) intentionally adds
'enabled' to normalized slot dicts. The two endpoint tests asserted the
exact key set {provider, model} — convert them to subset + round-trip
contracts so optional slot keys (enabled, reasoning_effort, max_tokens)
don't break them again.
Follow-up for salvaged PR #59753 rebased over the per-slot
reasoning_effort feature: _clean_slot now round-trips reasoning_effort
AND enabled together; add a normalize→normalize regression test, update
the validate/normalize agreement contract for the canonical enabled
default, restore the desktop per-slot toggle test on the current
autosave editor, and map oppenheimor's contributor email.
Frontend consumers for the events added by PR #59646: the TUI shows a
replace-in-place 'MoA: refs k/n' activity line (swapped for 'MoA:
aggregating…' on the aggregator phase), and desktop streams '◇ MoA refs
k/n' lines into the reasoning disclosure, self-cleaned by the first
moa.reference block.
Adds per-reference progress events and a phase-transition marker to the
MoA display pipeline so TUI / CLI / desktop surfaces can render a status
bar like `MOA: 2/3 refs done` and surface which phase (reference vs
aggregator) is currently active.
- `moa.progress` — fired once per reference completion with
`refs_done`, `refs_total`, and the source label
- `moa.phase` — fired on phase transitions (currently the single
`phase="aggregator"` transition once the fan-out
finishes)
Plumbed through the existing `reference_callback` →
`tool_progress_callback` → gateway path; no new UI surface. The legacy
`moa.reference` / `moa.aggregating` events are unchanged for backwards
compatibility.
AI-assisted fix by https://github.com/SquabbyZ/peaks-loop
Salvaged from PR #59743. Original author email was malformed
(sr@samirusani, not resolvable to a GitHub account), so the commit is
re-authored with credit via trailer.
Co-authored-by: Sami Rusani <samrusani@users.noreply.github.com>
Maintainer review (hermes-sweeper) on this PR found the fix was
incomplete: two paths still hid the MoA reference panel under
thinking: hidden.
1. thinking.tsx: the mount useState correctly seeds openThinking from
(visible.thinking === 'expanded' || reasoningAlwaysVisible), but the
re-sync effect on [visible] fires after the FIRST render too, not
just later updates, and lacks the reasoningAlwaysVisible OR — so it
immediately collapsed a just-opened MoA panel right after mount.
Skip only the effect's very first run (a ref flag); every later
visible change still re-syncs without the override, preserving the
documented no-OR-at-effect-time contract (manual collapse sticks).
2. useMainApp.ts: showProgressArea's streamSegments predicate gated
thinking content on thinkingPanelVisible alone, so an MoA reference
segment (segment.isMoaReference, same flag messageLine.tsx's
shouldShowThinkingTrail already honors per #64657) never kept the
live progress area up when thinking was hidden — StreamingAssistant
then returned early before MessageLine was ever reached. Added the
same override.
Added tests/thinkingMoaReferenceVisibility.test.tsx: mounts ToolTrail
with reasoningAlwaysVisible + sections.thinking: hidden, awaits queued
effects, and asserts the chevron is still open (▾, not ▸) once they
settle.
Validation:
npx vitest run src/__tests__/thinkingMoaReferenceVisibility.test.tsx
-> 1 passed
Fail-before: reverting only the thinking.tsx ref-guard reproduces the
exact regression -- the same test's frame capture shows the panel
open on first paint then collapsing to ▸ once the effect fires, and
the 'not.toContain(▸)' assertion fails as expected.
npx vitest run (full ui-tui suite): 1115 passed, 8 failed -- all 8
pre-existing and unrelated (terminalSetup/terminalParity/editor
resolution env-path tests), confirmed by running them in isolation
with the same result regardless of this diff.
npx tsc --noEmit: clean.
npx eslint src/components/thinking.tsx src/app/useMainApp.ts: clean.
Every moa.reference gateway event stores its labelled reference-model
output in a Msg's generic `thinking` field (turnController's
recordMoaReference), which messageLine.tsx and the ToolTrail component gate
on `display.sections.thinking`'s resolved mode. When that mode resolves to
`hidden`, MoA reference blocks were suppressed along with ordinary model
reasoning — even though (per #53855) references are the mixture-of-agents
process the user explicitly opted into, not private reasoning, and should
stay visible regardless of the thinking-section setting.
Adds Msg.isMoaReference (set by recordMoaReference), a shouldShowThinkingTrail
helper mirroring the existing shouldShowResponseSeparator pattern, and a
reasoningAlwaysVisible prop threaded into ToolTrail to bypass the two
suppression gates (the trail-wrapper return-null check and the
allHidden/panel-push checks) plus the panel's initial open state and the
shift-click expand-all gesture, so a MoA reference panel is not just present
in the tree but actually visible and openable on first paint.
Fixes#64657
Every moa.reference event called appendReasoningDelta(..., replace=true),
which wipes ALL existing reasoning-type message parts and seeds exactly one
new part. With two or more MoA reference models, each later reference
erased the reasoning disclosure built by earlier references, so only the
last advisor's output ever stayed visible instead of one labelled block per
reference (contradicting the multi-reference visibility behavior from
#53855).
Only the first reference (index <= 1, or missing) now replaces — preserving
the original "clear stale reasoning from before this turn" behavior. Every
later reference accumulates via the existing queue-then-flush path instead,
applied immediately since each reference arrives as one complete block
rather than incremental tokens.
Fixes#64658
The extracted findGitBash builds Windows-style candidate paths, but the
vitest suite (and any POSIX CI host) runs with posix path.join, which
mangles 'C:\Program Files' + segments into slash-joined paths and broke
the invalid-override fallback test. Use path.win32.join explicitly so
candidate construction is host-independent.
Port the HERMES_GIT_BASH_PATH env var check from main.cjs to main.ts
after the TS conversion. Also extract findGitBash to a dedicated module
for testability and add focused regression tests for override precedence
and invalid-override fallback.
Regression tests for the #47971 salvage: the LSP language-server spawn
must pass windows_hide_flags() creationflags while keeping PIPE stdio
and start_new_session, and the npm/go LSP auto-installer subprocess.run
calls must carry the same hide flags with DEVNULL stdin and
capture_output intact.
Salvaged from PR #47971 (LSP subset). On Windows, .cmd-wrapped language
servers (e.g. pyright-langserver.CMD launched via cmd.exe /c) and the
npm/go/pip LSP auto-installers spawn without CREATE_NO_WINDOW, so a
console window flashes whenever the spawn happens under a console-less
parent — e.g. a VS Code/Zed extension host running the ACP adapter.
- agent/lsp/client.py::_spawn: pass creationflags=windows_hide_flags()
to the language-server asyncio subprocess (inert 0 on POSIX;
start_new_session is kept — it is POSIX-only and ignored on Windows).
- agent/lsp/install.py: same flags on the npm and go installer
subprocess.run calls. The pip path goes through
hermes_cli.tools_config._pip_install, which already hides its windows.
Adapted from the PR's hand-rolled _NO_WINDOW constant to the repo's
hermes_cli._subprocess_compat.windows_hide_flags() convention.
The Windows /restart watcher's outer Popen spawns the watcher with
windows_detach_popen_kwargs() (which carries CREATE_BREAKAWAY_FROM_JOB),
but a restrictive parent job object can reject that bit with OSError and
the current call has no retry. Preserve the current watcher
implementation and add a focused breakaway-denied fallback.
Preserved from current main: watcher_python / pythonw.exe selection, the
str(restart_after_s) deadline, the scrubbed watcher_env, the intentional
no-breakaway inline respawn, and the entire POSIX setsid/bash path.
- primary keeps **windows_detach_popen_kwargs()
- on OSError, retry the same argv/env with
creationflags=windows_detach_flags_without_breakaway()
- on dual failure, log a definitive, path-safe warning (interpreter
basename + numeric winerror/errno only) and return without crashing
Replace the superseded breakaway-first inline design and its AST tests
with focused behavioral coverage that drives the real coroutine with a
mocked subprocess.Popen (retry, argv/env/DEVNULL preservation, POSIX
single-session kwarg, no-breakaway inline respawn, secret-safe logging).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds moa.privacy_filter ('' | display | full, default off — issue #59959):
- display: redact user-visible surfaces only (reference blocks emitted to
the UI + saved MoA trace records, including per-advisor full input/output
and the aggregator-input copy); the aggregator sees raw advisor text so
synthesis quality is unaffected.
- full: additionally redact the advisor text injected into the aggregator
prompt, on both the persistent facade path and the one-shot /moa
synthesis path (the issue's literal ask). Legacy boolean true maps here.
Secret/credential shapes (API-key prefixes, JWTs, private keys, DB
connection strings) are delegated to the central redactor
(agent.redact.redact_sensitive_text, force=True + code_file=True); the MoA
filter adds only email and clearly delimited phone-number patterns. No
bare 10-digit matching: line numbers, timestamps, epoch values, git SHAs,
IPs, versions, and source-code assignments in code-review-shaped advisory
text pass through byte-identical. The reference cache always holds raw
text — redaction happens at each consuming surface, so a mid-session mode
change never leaks or double-redacts.
Reworked from PR #60463: replaced its hand-rolled pattern list (which
matched bare digit runs and re-implemented key shapes) with central-
redactor reuse + safe patterns, and split the single boolean into
display/full modes. Credited for the feature framing.
Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com>
Extends the fanout enum with 'every_n:<N>' (N >= 2): advisors run on the
first iteration of each user turn and every Nth tool iteration after it;
off-cadence iterations REUSE the cached guidance from the last on-cadence
run via the same cache mechanism the user_turn fanout uses, so the
aggregator still gets advice on every step. The cadence counter is scoped
per user turn (resets on a new user message) and only advances when the
advisory state actually changes, so streaming retries never consume a
cadence slot. Mapping form {mode: every_n, n: N} normalizes to the
canonical string. Unknown/degenerate values fall back to per_iteration.
Addresses issue #63393 (advisor fan-out multiplies turn latency/cost by
the tool-iteration count). Redesigned from PR #63448: the submitted shape
skipped references entirely on off-cadence iterations (aggregator ran
advice-less); this version keeps the last advice in play, credited for
the idea and cadence framing.
Config-gated, default-off (default fanout remains per_iteration).
Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com>
Adds test_moa_gemini_aggregator_sanitize_uses_real_model: drives a full MoA
tool-call turn (virtual-provider mode) with a Gemini aggregator and asserts
the strict-API sanitize pass is invoked with the resolved aggregator model
(gemini-3-pro-preview), never the virtual preset name once a slot is
resolved — the exact path that stripped extra_content/thought_signature and
made Gemini aggregators 400 (#65092).
Writing the test surfaced a gap in the salvaged #66212 fix: in virtual-
provider MoA mode (provider=moa, no moa_config threaded through
run_conversation) the conversation-loop branch never fired because it only
consulted moa_config. Extend it to fall back to the facade's
last_aggregator_slot — the same source the handle_max_iterations fix uses —
so both MoA entry modes resolve the real aggregator model.
Also adds the contributors/emails mapping for the #15676 credit base.
Follow-ups to the salvaged core of #60293:
- Gate the x-initiator header on _normalize_aux_provider() instead of a
literal 'copilot' string compare, so slot configs spelled github /
github-copilot / github-models / copilot-acp / mixed case all get the
user-turn attribution.
- Thread extra_headers through _retry_same_provider_sync/_async so the
credential-refresh and pool-rotation retry rebuilds don't silently drop
the header (the rebuilt kwargs previously started from scratch).
- Add a transport-boundary test asserting the header reaches the SDK
client's create() kwargs (no call_llm mocking), an alias-spelling
matrix test, and a retry-rebuild preservation test.
Follow-up to the salvaged core of #53802: a naive MoAClient(preset) rebuild
restores a working facade but silently drops the reference_callback relay
wired in agent_init, so moa.reference / moa.aggregating display events stop
reaching every frontend for the rest of the session.
Introduce agent.moa_loop.build_moa_facade(agent, preset) as the single
construction point for the MoA facade and use it at:
- initial client construction (agent_init.py)
- turn-start fallback restore (restore_primary_runtime)
- transient transport recovery (try_recover_primary_transport — previously
fell through to _create_openai_client with MoA's empty client_kwargs and
died with 'api_key client option must be set')
- mid-session model switches (switch_model)
The relay reads agent.tool_progress_callback at emit time, so callbacks
attached after construction are picked up automatically.
Adds test_moa_restored_facade_still_emits_reference_events covering event
delivery through a restored facade.
When Hermes fails over from a non-Gemini provider (xAI, Anthropic, etc.) to
Gemini mid-conversation, the existing assistant tool_calls in history carry
no Gemini ``extra_content.google.thought_signature`` (the originating provider
never emits one). The native adapter's ``_translate_tool_call_to_gemini``
omitted ``thoughtSignature`` entirely in that case, so Gemini 3 thinking
models rejected every replayed turn with::
HTTP 400 INVALID_ARGUMENT
Function call is missing a thought_signature in functionCall parts.
Additional data, function call default_api:<tool_name>, position N.
The Cloud Code Assist sibling adapter already handles this exact case by
emitting a sentinel ``"skip_thought_signature_validator"`` (see
``agent/gemini_cloudcode_adapter.py:106``, originally added in #11270 and
documented as matching ``opencode-gemini-auth``'s approach). This change
mirrors that fallback in the native adapter so the two paths behave
identically when replaying cross-provider history.
Verified live against ``generativelanguage.googleapis.com/v1beta`` with
``gemini-3-pro-preview``: synthetic 2-turn conversation with no real
``thoughtSignature`` returns 400 without the sentinel and 200 with it.
Test added: ``test_build_native_request_emits_sentinel_for_cross_provider_tool_call``.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When MoA mode is active with a Gemini model as the aggregator,
agent.model holds the virtual preset name (e.g. "closed"), not the
actual aggregator model name. The _sanitize_tool_calls_for_strict_api
call uses agent.model to decide whether to keep extra_content
(thought_signature) on tool_calls — since "closed" doesn't contain
"gemini", the thought_signature is stripped and the Gemini aggregator
rejects the next request with HTTP 400 (INVALID_ARGUMENT):
"Function call is missing a thought_signature in functionCall parts."
Fix: resolve the actual aggregator model name from moa_config
(conversation_loop) or last_aggregator_slot (chat_completion_helpers)
and pass it to _sanitize_tool_calls_for_strict_api so the
_model_consumes_thought_signature check sees the real Gemini model.
Closes#65092
Bare ContextCompressor.__new__ doubles (test_compress_focus,
cross_session_guard, image_tokens, pre_compress_memory_context) skip
__init__ and lack the attribute — the documented compression-path
test-double pitfall. Guard with getattr default 1 + int type pin
(bool excluded).
Follow-up fixes on top of the salvaged #22566 mechanism:
- N-collector now counts only REAL actionable user turns via
_is_actionable_user_turn + _is_synthetic_compression_user_turn —
the same filter pair _find_last_user_message_idx uses post-#69291.
The contributor's bare role=='user' + _is_context_summary_content
check let blank platform echoes and continuation/todo rows consume
N slots, silently degrading the guarantee.
- Default flipped 3 -> 1 (behavior-preserving): a default of 3 was
measured to change the tail cut on transcripts whose budget covers
only the last turn. min_tail_user_messages=1 delegates to the
existing single-user anchor; N>1 is opt-in, and the call site is
gated so the default path is byte-identical to main.
- Hardened config parse in agent_init (bool rejected, fractional
floats rejected, floor 1) matching the max_attempts parser shape.
- Wired the recurring external-PR config gaps: hermes_cli/config.py
DEFAULT_CONFIG + cli-config.yaml.example (PR only had cli.py).
- Regression tests: blank echoes / synthetic rows don't count toward
N; tool-call/result pairs never split by the N-boundary (no-orphan
both directions); N-guarantee wins over tail_token_budget and the
_MAX_TAIL_MESSAGE_FLOOR (floor is a minimum, not a cap); default
parity pin; DEFAULT_CONFIG pin.
Add _ensure_last_n_user_messages_in_tail to guarantee the last N user
messages survive compression in the uncompressed tail, with surrounding
assistant/tool context preserved.
- Add min_tail_user_messages parameter (default 3) to ContextCompressor
- New _ensure_last_n_user_messages_in_tail method generalizes single-user protection
- Skip context-summary handoff banners when counting user messages
- User messages are clean boundaries — skip _align_boundary_backward
- Wire through cli.py, agent_init.py, and gateway cache busting keys
Config:
compression:
min_tail_user_messages: 3
Co-Authored-By: Claude <noreply@anthropic.com>
_collect_ghosted_skill_names() covers both ghost-skill shapes in the
compressed middle window: rows already demoted to a [SKILL_PRUNED: ...]
marker AND raw skill_view bodies (> _SKILL_VIEW_PRUNE_MIN_CHARS) that
survived Phase-1 inside an earlier protected tail and then aged into the
compression window — the summarizer paraphrases those instructions away
too. Shared threshold constant between the emit site and the scan.
Pinned by a live-probe-shaped test (real compress(), mocked aux LLM).
21 tests pinning the salvaged #44166 behavior:
- marker emit + extractor round trip (patterns adapted from PR #32375
by @LeonSGP43, with credit)
- no-duplicate re-injection when the canonical marker survived (the
original PR's presence-check defect)
- Phase-1 protection for just-loaded / user-referenced skills, and the
Pass-4 pressure override that keeps #61932 fixed
- deterministic marker survival through a REAL compress() with a mocked
aux LLM: drop → re-injected, keep → not duplicated, static-fallback
path, iterative re-compression via rehydrated handoff
- markers never classify as handoff content (classify_summary_content /
_strip_context_summary_handoff_message untouched)
- SKILLS_GUIDANCE Skill Safety Rule renders with real newlines
Salvage rework of PR #44166 (@dolphin-creator) onto current main:
- ONE canonical prune marker: _skill_pruned_marker(name) builds
'[SKILL_PRUNED: ... reload with skill_view(name='X')]'; both emit
sites and the survival presence check use the same string, fixing the
original PR's defect where the emitted marker was '[SKILL_PRUNED:'
but the presence check looked for '[SKILL_PRUNED]' (re-injection
duplicated markers that had survived).
- Phase-1 prune (_prune_old_tool_results) now threads a protected-skill
set: skills whose skill_view call is within the last 10 messages, in
the protected tail, or named in a tail user message keep their full
bodies. Pass-4 pressure demotion deliberately overrides the guard so
the #61932 dead-end shape cannot return.
- P2 deterministic marker survival: skill names are extracted from the
summarizer INPUT (and the previous summary) before the aux LLM call
and any dropped canonical markers are re-injected afterward under a
'## Pruned Skills' section — routed through _redact_compaction_text,
appended to the summary body only (never in front of SUMMARY_PREFIX
or scaffolding start-of-content markers; classify_summary_content is
unaffected). Same treatment on the static fallback path, re-applied
after its size cap since truncation cuts exactly where markers land.
- Summarizer prompt gains a '## Pruned Skills' copy-verbatim section.
Fixes#32106.
Surgical reapply of the marker-alignment and dedup-guidance halves of
PR #44166 commits 52341f6ca3 / 3d8a31432d / ae07412e4b onto current main:
- the [SKILL_PRUNED: ...] marker embeds the exact reload call
skill_view(name='<skill>') so the model can act without guessing
- SKILLS_GUIDANCE Skill Safety Rule gains rule 4 (DEDUP): after one
reload, remaining markers for the same skill are historical artifacts
Fixes#32106 (part).
Community verification of #56688 (zmack12344321) found two follow-up gaps
that kept Vertex invisible in the /model menu even after registry
registration:
1. hermes_cli/model_switch.py: list_authenticated_providers() had a
credential gate hard-coded to API keys (with an aws_sdk special case
only) — add a vertex branch using has_vertex_credentials(), mirroring
the aws_sdk shape.
2. hermes_cli/models.py: Vertex's OpenAI-compatible endpoint has no
/models listing route, so without a curated _PROVIDER_MODELS entry the
picker only ever showed the current model — add a Gemini curated list.
Follow-up to #56688.
The Vertex AI provider (added same-day, commit c73e74386) was never added to
either of the two provider registries that agent/auxiliary_client.py and the
MoA slot-resolution chain depend on, breaking Vertex outside the main
conversation loop:
1. hermes_cli/auth.py::PROVIDER_REGISTRY had no "vertex" entry. The
plugin-auto-extend loop that normally fills gaps explicitly skips
non-api_key auth types (`if _pp.auth_type != "api_key": continue`), and
Vertex was never hand-declared like "bedrock" is. Because
resolve_provider_client() in agent/auxiliary_client.py gates everything
on `pconfig = PROVIDER_REGISTRY.get(provider)` and returns (None, None)
immediately when pconfig is None, its `elif pconfig.auth_type == "vertex"`
branch was permanently dead code — every auxiliary Vertex call (vision,
title generation, reflection, context compression, MoA reference/
aggregator slots) failed outright, not just a MoA-specific edge case.
2. hermes_cli/providers.py::HERMES_OVERLAYS also had no "vertex" entry, so
hermes_cli.providers.get_provider("vertex") returned None. This backs
_preserve_provider_with_base_url() in agent/auxiliary_client.py, which a
MoA slot's resolved (base_url, api_key) pair needs to keep its "vertex"
identity instead of silently collapsing to "custom" — losing the
identity _refresh_provider_credentials() needs to re-mint an expired
OAuth2 token (~1h lifetime) on a 401, and permanently breaking every
subsequent call in that MoA preset for the rest of the session.
Fix mirrors the existing "bedrock"/aws_sdk entries in both registries
exactly, plus adds a "vertex" branch to _refresh_provider_credentials() (it
had branches for openai-codex/nous/anthropic/xai-oauth but not vertex,
so a 401 fell through to `return False` without evicting the stale cached
client).
- hermes_cli/auth.py: hand-declared vertex ProviderConfig(auth_type="vertex")
in PROVIDER_REGISTRY, matching bedrock's shape.
- hermes_cli/providers.py: vertex HermesOverlay(auth_type="vertex") in
HERMES_OVERLAYS + "Google Vertex AI" label override.
- agent/auxiliary_client.py: vertex branch in _refresh_provider_credentials
that re-mints the token via get_vertex_config() and evicts the stale
cached client.
- 8 new regression tests across tests/hermes_cli/test_vertex_provider.py and
tests/agent/test_auxiliary_client.py: registry membership, end-to-end
resolve_provider_client("vertex", ...) building a working client (proving
the previously-dead branch is now reachable), and the 401-refresh/cache-
eviction path.
MoaConfigPayload does not declare save_traces or trace_dir, so
set_moa_models() overwrites cfg["moa"] with a dict that lacks these
hand-edited keys. Use dict.update() to merge instead of replace.
Fixes#58819
When reference_models is stored as a JSON string (e.g. from hermes moa
configure or hand-edited config.yaml), _normalize_preset silently
falls back to hardcoded defaults because the string fails both
isinstance(x, list) and isinstance(x, dict) checks.
Add json.loads() parsing before the type checks so both formats work.
MoA (Mixture of Agents) is a legitimate internal provider used by
Diagnosis presets and multi-model aggregation. When a MoA preset sets
model.provider to 'moa', hermes doctor incorrectly reports it as
'unrecognised' and suggests changing it, which would break the MoA setup.
Add 'moa' to the known_providers set alongside 'openrouter', 'custom',
and 'auto' so doctor recognises it as valid.
Fixes#58759
The existing tests cover _normalize_moa_model() in isolation and a local
precedence expression, but not the __init__ wiring itself. Add two
init-level regression tests: constructing HermesCLI(model='moa:strategy')
strips the prefix to model='strategy' and forces requested_provider='moa',
and the moa: prefix wins over an explicit --provider. Both fail if the
override is dropped from the requested_provider resolution.
Refs #56828
hermes chat -Q -m moa:strategy failed with 'model moa:strategy is not
supported' (HTTP 401/400): the raw model string was passed straight to
the real provider. The MoA virtual provider only got wired up through the
interactive /moa command and the model picker, never through the -Q
one-shot startup path.
resolve_runtime_provider already handles requested_provider == 'moa', and
agent_init builds the MoAClient off provider == 'moa' (surface-agnostic).
The only gap was mapping the moa:<preset> model string to that provider.
Add _normalize_moa_model() and apply it in HermesCLI.__init__ before
provider resolution: a moa:<preset> model sets requested_provider='moa'
and model=<preset>, so the existing MoA path runs in non-interactive mode
too. The moa: prefix wins over an explicit --provider (previously
--provider deepseek -m moa:strategy silently dropped MoA).
Fixes#56828
Follow-ups on top of the cherry-picked #27748 mechanism:
- move the cap constant to module level with full rationale comment
(class attribute aliases it so subclasses/tests can override)
- bound the iterative-update path too: the PREVIOUS SUMMARY block is
passed through _bound_summary_input so a pathological rehydrated
handoff cannot blow up the prompt (previous summary + new turns each
capped)
- extra regression tests: byte-identical small-input passthrough
(identity), direct bound+marker unit check, bound-after-per-message-
truncation shape (hundreds of under-_CONTENT_MAX turns), iterative
path bounded, marker vs classify_summary_content non-collision
- contributor email mapping for @robgfl45
Follow-ups on top of the cherry-picked #62644 mechanism, porting it to
current main and closing the salvage-review requirements:
- proactive_prune_min_reclaim_tokens (default 4096): a prune only COMMITS
when it reclaims a meaningful token batch, measured on the pruned output.
A committed prune rewrites already-sent history and invalidates the
provider prompt-cache prefix; this hysteresis gate keeps those breaks
episodic/amortized (like a compression boundary) instead of firing every
tool iteration. 0 disables the gate. (Design point credited to the
#62389 review cycle's prune_minimum_tokens.)
- Standard no-op caller contract: every skip path returns the INPUT list
object; the loop commits only on 'result is not messages' + non-zero count.
- Loop call is getattr+callable guarded (plugin engines predating the hook,
SimpleNamespace test doubles) and exception-swallowed at debug level.
- Config parse follows the compression.max_attempts hardened semantics:
booleans rejected, fractional floats rejected, integral floats/numeric
strings accepted; negative trigger = disabled.
- cli-config.yaml.example documented (all three keys) and gateway
_CACHE_BUSTING_CONFIG_KEYS extended so hot-reload rebuilds the agent.
- Tests: min-reclaim gate both directions, input-object no-op contract,
no-orphan tool_call_id pairing in BOTH directions (#69830 pin rule),
default-off zero-behavior-change pin, config parse seam, and behavioral
loop-wiring tests (consulted/commit/no-op/absent-method/raising).
The phase-1 tool-result prune only runs inside compress(), which fires
near 50% of the context window, so it never triggers on large-window
models; old tool outputs then ride in history and are re-sent every turn.
Add prune_tool_results_only(): the same no-LLM prune on a separate, low
proactive_prune_tokens trigger, run as an elif to the compression branch.
Opt-in (default 0), protects the recent tail by message count.
Add the method to the ContextEngine base as a no-op default so pluggable
engines inherit it safely (the post-tool-call path never AttributeErrors on
a non-built-in engine); the built-in compressor supplies the real prune.
Register both keys under the top-level compression config with defaults and
document them.
Follow-up to the #54805 cherry-pick (skill guard rules):
- turn_context.py: snapshot_preflight_display_tokens gets a
getattr+callable guard (SimpleNamespace compressor doubles / plugin
context engines lack the ContextCompressor-only method) and the
snapshot value is type-pinned to a real int (bool excluded) before
arming the rollback — MagicMock compressors return truthy Mocks.
- turn_finalizer.py: interrupted pinned 'is True', the
_turn_received_provider_response read pinned 'is not True' (MagicMock
auto-attrs are truthy), and the compressor rollback method call gets a
getattr+callable guard. Rollback stays display-only: it never touches
_ineffective_compression_count or any durable guard, and preserves the
-1 post-compaction sentinel.
Interrupted turns can seed a speculative display token count before the provider receives the request. Restore that display-only seed when interruption wins the race, while preserving completed post-compaction state and treating a successful provider response independently of optional usage metadata.
Constraint: #54776 remains reproducible on current main, while review #4702305384 identifies anti-thrashing rollback as stale and usage receipt as an unreliable response-completion signal.
Rejected: Restore anti-thrashing counters from a preflight snapshot | current main derives their verdict from real provider usage after a completed compaction boundary.
Confidence: high
Scope-risk: narrow
Directive: Keep interrupted preflight rollback display-only, and never infer provider completion from the presence of usage metadata.
Tested: ./.venv/bin/python -m pytest -q tests/run_agent/test_413_compression.py (29 passed); turn-finalizer/conversation-loop tests (31 passed); context-compressor targeted tests (12 passed); infinite-compaction targeted tests (3 passed); ruff; git diff --check.
Not-tested: End-to-end interactive interrupt through CLI or gateway transport.
Gateway half of the #49874 salvage: pass compression_deferred through
both _run_agent_inner result dicts and guard the compression-exhausted
auto-reset block with it — a lock-contended defer keeps the session
intact (the concurrent compressor is actively shrinking it) instead of
wiping it via reset_session.
Regression tests:
- tests/run_agent/test_compression_lock_defer.py — provider-mock 413 and
400-overflow turns whose compression pass lost the lock end as
compression_deferred (failed=False, no compression_exhausted); flag
unset keeps the terminal exhaustion path byte-identical; type-pin
tests vs MagicMock agents and junk flag values; cap=1 e2e proving the
refunded pre-API defer leaves the budget for the provider-proven
413 retry.
- tests/agent/test_preflight_lock_defer.py — a lock-skipped preflight
pass stops the loop WITHOUT arming preflight_compression_blocked;
plain no-op still arms it; MagicMock junk does not defer.
- tests/gateway/test_compression_deferred_soft_result.py — AST pin that
the deferred branch guards the auto-reset chain and performs no
session mutation (mirrors test_35809_auto_reset_clean_context.py).
A lock-loser compression pass returns its input unchanged, which the
automatic compression sites misread as 'cannot compress further': the
preflight loop armed the insufficient-progress blocker, the pre-API gate
burned a shared attempt, and a lock-contended 413/overflow retried into
the attempt cap and returned compression_exhausted — which the gateway
answers with a full session auto-reset (#9893/#35809). A temporary
concurrent-compression defer wiped the session.
Consume the landed #69870 lock-skip signal on every automatic path
(preflight in turn_context, pre-API pressure gate, 413 handler, overflow
handler, post-tool compaction): when a pass no-ops AND the type-pinned
lock-skip flag is set, refund the attempt (never count it toward the cap
or the insufficient-progress blocker), and when the turn cannot proceed
(provider already proved the request does not fit) end it with a soft
compression_deferred result — distinct from compression_exhausted — so
the gateway keeps the session intact and the next message retries after
the concurrent compressor finishes.
The new compression_skipped_due_to_lock() reader is type-pinned
(is True or isinstance(str)) per the MagicMock auto-attribute rule, and
compress_context() now also clears the signal at the very top of every
attempt (per-attempt state rule, #58629/#69853) so a stale value can
never make a later breaker/codex no-op look like lock contention.
Salvaged from PR #49874; rebuilt on main's #69870
_compression_skipped_due_to_lock signal instead of the PR's parallel
_compression_deferred_by_lock triple.
Hardening on top of the salvaged dead-PID lease reclamation from PR #65775
(@the3asic):
- Probe via psutil.pid_exists (hard dependency; CONTRIBUTING.md critical
rule #1) with the contributor's os.kill(pid, 0) POSIX probe retained
only as a scaffold-phase fallback when psutil is missing.
- Same-process holders (pid == os.getpid()) are never probed and never
self-reclaimed — another thread's live lease is owned by the lease
refresher/release path.
- Any probe doubt (exceptions, permission errors) conservatively keeps
the lease until normal TTL expiry; Windows stays TTL-only.
- Tests: psutil-first dead-pid reclaim (probe call pinned), os.kill
fallback path, probe-doubt keeps lease, same-pid no self-reclaim,
legacy holder + Windows paths assert NO probe via either API.
compress_trajectory (and _async) replaced the compressible middle region with
a [CONTEXT SUMMARY] turn without checking that the region is actually larger
than the summary. When a large protected system prompt dominates the budget,
the compressible middle can be tiny; replacing e.g. a 2-token middle with a
~60-token summary GROWS the trajectory (tokens_saved negative), marks it
was_compressed, and still spends a summarization call — the opposite of the
intent, on exactly the hard over-budget cases.
Add a net-savings guard mirroring the code's own comment (net_savings =
region_tokens - summary_target_tokens): if the safely-compressible region is
no larger than summary_target_tokens, return the trajectory unchanged. Applied
to both the sync and async paths. Add sync+async regression tests.
MoA reference_max_tokens is preset-level — one cap for all reference
models. When mixing a verbose model with a terse one, a single cap is
either too tight for the terse model or too loose for the verbose one.
Now each reference slot can optionally carry its own max_tokens:
reference_models:
- provider: openrouter
model: deepseek/deepseek-v4-pro
max_tokens: *** # per-slot cap, overrides preset-level
- provider: openai-codex
model: gpt-5.5
# no max_tokens → falls back to preset-level reference_max_tokens
_clean_slot (moa_config.py) preserves an optional max_tokens field on
the slot dict, coerced via _coerce_int_or_none. _run_reference
(moa_loop.py) reads slot-level max_tokens first, falling back to the
preset-level cap passed by the caller. Slots without the field are
unaffected — backward compatible.
Type hints on slot-handling functions updated from dict[str, str] to
dict[str, Any] to reflect the now-heterogeneous slot shape.
Requested in review: builder-level assertions that the gemini-native
branch forwards max_tokens (provider names and the native
generativelanguage.googleapis.com base_url, max_tokens=600), plus a
control showing gemini models on OpenAI-compatible endpoints — including
Gemini's own /openai compatibility endpoint — keep the existing omission
behavior (#34530).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
_build_call_kwargs omitted max_tokens for every provider except
anthropic-compat endpoints and NVIDIA NIM. Gemini's native
generateContent maps max_tokens -> maxOutputTokens and, when it is
omitted, applies a fixed 65,535-token ceiling (not "the model's full
budget"), so dropping the value made MoA's reference_max_tokens a
silent no-op for gemini advisors — they ran effectively uncapped
(observed ~2900 output tokens against a configured cap of 600),
inflating per-turn MoA latency.
Forward max_tokens for the gemini-native path (provider name or native
base_url). Gemini supports maxOutputTokens, so the cap is safe here;
providers that reject max_tokens (Copilot, GPT-5 max_completion_tokens,
ZAI vision) are unaffected — they still omit it as before.
Per review feedback from teknium1: reference_max_tokens is an advisors-only
contract. The aggregator is the acting model and must not be capped by the
reference budget. Changed _is_moa from startswith('moa_') to exact match on
'moa_reference'. Added regression test proving aggregator does NOT receive
max_tokens.
Copilot review pointed out that hardcoding kwargs['max_tokens'] would
400 on models requiring max_completion_tokens (GPT-5 family, Copilot).
The existing auxiliary_max_tokens_param() helper already selects the
correct parameter name per model — use it instead of hardcoding.
Test updated to parametrize expected_key so the Copilot gpt-5.5 case
correctly asserts max_completion_tokens instead of max_tokens.
Addresses Copilot review comments on both files.
PR #56756 added reference_max_tokens to cap MoA advisor output and cut
turn latency. The value is correctly threaded through five layers of MoA
code (moa_config → conversation_loop → aggregate_moa_context →
_run_references_parallel → _run_reference → call_llm(task='moa_reference',
max_tokens=800, ...)).
However, _build_call_kwargs() in auxiliary_client.py silently drops
max_tokens for all OpenAI-compatible providers (PR #34845, which fixed
endpoints and NVIDIA NIM keep it. This means reference_max_tokens never
reached the API for the vast majority of providers.
The bug affects every OpenAI-compatible MoA reference/aggregator slot:
Z.AI (coding plan), OpenRouter, OpenAI, GitHub Copilot, and local
providers. Only Anthropic-compat endpoints (MiniMax, /anthropic URLs)
worked — by coincidence, not MoA-aware design.
Fix: thread the 'task' parameter through all six _build_call_kwargs()
call sites. When task starts with 'moa_', max_tokens is always included
in the request kwargs regardless of provider. Non-MoA auxiliary tasks
(compression, titles, vision, etc.) keep PR #34845 behavior unchanged.
Verified end-to-end:
- Z.AI GLM-5.2 with max_tokens=50 → returned exactly 50 tokens
- Z.AI GLM-5.2 with max_tokens=20 → returned exactly 20 tokens
- Z.AI GLM-5.2 uncapped → returned 315 tokens
- 7 new regression tests covering 4 providers, Anthropic wire, non-MoA
tasks, and prefix-matching boundary
- 288 auxiliary_client tests pass (was 281, +7 new), 84 MoA tests pass
- Zero regressions
aggregate_moa_context's single max_tokens parameter was applied to
both the reference fan-out (_run_references_parallel) and the
aggregator's own synthesis call_llm. #53580 explicitly removed a
hardcoded cap from the aggregator call because it truncated long
aggregator syntheses; #56756 (reference_max_tokens, added to speed up
the advisor fan-out) reintroduced the same shared cap by passing it to
both calls, silently regressing #53580's fix.
Rename the parameter to reference_max_tokens (matching the caller's
own moa_config key) and stop forwarding it to the aggregator's
call_llm invocation, which now always runs uncapped as intended.
Every npm workspace package now defines check:* scripts (check:unit,
check:lint, check:bundle, check:typecheck, etc.) that fan out to
separate matrix runners in CI. The check umbrella script chains all
shards for local dev.
The matrix discovery in the workspaces job queries npm workspaces,
finds check:* scripts (in package.json insertion order), falls back to
check when none exist, and emits an include matrix. No hardcoded
package names — the workflow is fully auto-derived from workspace
metadata.
Previously every package ran a single check script on one worker, and
the fix step (lint:fix + prettier) ran as a separate CI step with
special-cased run_fix gating to avoid running on every shard. Now that
lint is just another check:lint shard, the run_fix field and the fix
step are gone entirely — lint runs in its own runner like everything
else.
An assistant message with an empty or whitespace-only text content block —
produced by context compression or certain tool-call flows — is rejected by
the Anthropic Messages API with HTTP 400 "text content blocks must contain
non-whitespace text". Because the blank block is stored in session history and
replayed verbatim every turn, the session is permanently wedged behind the same
400.
The Bedrock adapter already guards this via _safe_text() (#9486); the native
Anthropic path never got the same treatment. _sanitize_replay_block() rebuilt
text blocks with the raw stored text, and the _convert_assistant_message()
guard only caught a fully empty block list, not a list still containing a
whitespace-only text block.
Add a _safe_text() helper mirroring the Bedrock one and apply it at both points:
the ordered-blocks replay path and a final in-place walk of the converted
content list. Both are self-healing — sessions that stored blank blocks recover
on the next API call. Only text blocks are coerced; thinking/tool_use/image
blocks are untouched.
Fixes#69512
The renderer showed two assistant bubbles for one turn — a partial
streamed copy plus the clean final copy (#63679). A reload fixed it, so
the persisted transcript was correct; this was live reconciliation.
Root cause is in completeAssistantMessage (use-message-stream/index.ts).
message.interim fires for BOTH verify-on-stop candidates AND ordinary
tool-call turns (tui_gateway _load_interim_assistant_messages), and the
interim seal clears streamId + sets interimBoundaryPending. So at
message.complete the streamId fast-path is skipped and it enters the
fallback. There the settle-onto-interim branch was gated on
responsePreviewed — true ONLY for verify-on-stop. A normal tool-call
turn whose final text matched its sealed interim satisfied neither
settle branch and fell through to append a brand-new bubble. Two rows,
distinct ids, id-based dedup cannot collapse them → renders twice.
Fix: settle onto the sealed interim whenever the final CONTINUES it —
final == interim, final starts with interim (streamed + trailing delta),
or interim starts with final (streaming dropped characters). This is
gated on existing.interim so it only ever collapses a genuine sealed
interim, and a genuinely DIFFERENT final still appends as its own
bubble. responsePreviewed is retained as an OR so the verify-on-stop
continuation-budget case (final text rewritten, no shared prefix) still
settles as before.
The prior test that asserted a non-previewed identical interim+final
produced TWO bubbles was encoding the bug; rewritten to assert one.
Added coverage: prefix-extended non-previewed final collapses; a
genuinely different final still appends (no over-collapse).
Community analysis on the issue (seedSeenBubbleKeys / kind-guard) was
against a pre-refactor v0.7.0 bundle — those symbols no longer exist;
this is the current-code root cause.
Co-authored-by: SHL0MS <SHL0MS@users.noreply.github.com>
Gateway tests construct GatewayRunner via object.__new__ without __init__;
the new #51899 guard accessed self.config directly and crashed those
runners (AttributeError). Use getattr with None default — the guard
no-ops without config, matching the documented object.__new__ pitfall.
Note: #51899's commit was authored as a placeholder identity
(hermes-agent@example.invalid); it was re-stamped to @byshubham's
numeric noreply on this branch, so no mapping file is needed. Other
salvaged commits use GitHub noreply addresses.
Two related Slack delivery fixes for send_message text sends:
- Route Slack text delivery through _send_via_adapter so the live
in-process gateway adapter (multi-workspace aware, channel→client
mapping, adapter-side gates) is preferred, with the plugin's
_standalone_send as the out-of-process fallback — matching how the
media path already behaves.
- _standalone_send: SLACK_BOT_TOKEN can be a comma-separated list in
multi-workspace installs and slack_tokens.json carries OAuth
per-workspace tokens; the standalone Web-API path used to send the
literal comma-joined string, which Slack rejects as invalid_auth.
Try each token individually, retrying on token-scoped errors
(invalid_auth / not_in_channel / channel_not_found …) and stopping on
terminal ones. User-DM resolution (U…/W… targets) also tries each
token.
Adapted from #47547 by @replygirl — the original patched the legacy
tools/send_message_tool.py::_send_slack helper, which moved to the
Slack plugin's _standalone_send in #41112.
Salvaged from #47547
Previously, Slack sessions and the channel directory stored raw channel
IDs (e.g. C0ATFHY907L) as chat_name, making it impossible for operators
to identify channels in send_message listings or session data.
Changes:
- Add _channel_name_cache and _resolve_channel_name() to SlackAdapter
that calls conversations.info (channels) or users.info (DMs) with
in-memory caching to avoid repeated API calls
- Use resolved names in _handle_slack_message() build_source() calls
- Use cached names in _seed_assistant_thread_session()
- Enrich session-sourced entries in channel_directory._build_slack()
by cross-referencing API results and falling back to conversations.info
+ users.info for DMs and private channels not in the bot's scope
Fixes: channel directory and session origins now show readable names
like 'general' or 'John Doe' instead of 'C0xxx' / 'D0xxx'.
Follow-up to the #51899 pick, folding in the config-bridge half of the
competing PR #46925 (@bhanusharma, earliest submitter for the
ignored-channel gate):
- _apply_yaml_config: translate config.yaml slack.ignored_channels into
SLACK_IGNORED_CHANNELS (list or CSV), env-var-wins like every other
bridged Slack key.
- SlackAdapter._slack_ignored_channels / gateway.run's
_slack_ignored_channels_from_gateway_config: fall back to the
SLACK_IGNORED_CHANNELS env var when PlatformConfig.extra carries no
value, so top-level slack: blocks (which flow through the env bridge,
not extra) are honored at both the adapter and runner gates.
- conftest: force-clear SLACK_ALLOWED_CHANNELS / SLACK_IGNORED_CHANNELS /
SLACK_DISABLE_DMS between tests (config-loader side-effect leak class).
- Tests: env-bridge translation + precedence in test_config.py, env
fallback + extra-wins in test_slack_runner_ignored_channels.py.
Credit: #46925 by @bhanusharma proposed the same gate with the YAML→env
bridge; #51899 (picked as the base) carries the wider outbound/runner
coverage. Closes#46925 as consolidated here with first-submitter credit.
Add --long-description / --long-description-file to `hermes slack
manifest` so the generated app manifest can carry Slack's
display_information.long_description (175–4,000 characters), with
validation of the length bounds, mutual-exclusion with --slashes-only,
and UTF-8 file input. Also propagate the manifest command's exit status
through cmd_slack so validation failures reach the shell.
Squash of the two commits from PR #65256 — one commit per contributor
on this salvage branch.
Salvaged from #65256
PR #37303 added per-platform streaming defaults and the commit message
explicitly called out "Discord/Slack/etc. only have edit-based streaming
(repeated editMessage), which flickers and is noticeably jankier" — but
only discord.streaming=false was shipped. Slack uses the same edit-based
streaming mechanism and has the same flicker problem, yet it was left to
follow the global switch (default true when streaming is enabled).
Add "slack": {"streaming": False} to DEFAULT_CONFIG["display"]["platforms"]
alongside the Discord default. The same deep-merge semantics apply: a user
who explicitly sets display.platforms.slack.streaming: true keeps their
value unchanged. The dashboard schema gains a slack.streaming toggle
automatically since it is generated from DEFAULT_CONFIG.
Update test_per_platform_streaming_defaults.py to cover slack in all
existing assertions and rename the resolver test to reflect both platforms.
Hermes already identifies itself on outbound calls to model providers
and monitoring endpoints — see hermes_cli/models.py and
plugins/model-providers/gmi/__init__.py
("Attribution so GMI can identify traffic from Hermes Agent"). The
Slack adapter is the one outbound surface that doesn't carry the
same identifier, so HermesAgent-driven Slack traffic is
indistinguishable from any other Bolt-Python app at the Slack
platform layer.
This sets `user_agent_prefix=f"HermesAgent/{_HERMES_VERSION}"` on the
AsyncWebClient instances constructed in gateway/platforms/slack.py
and threads them through AsyncApp via its `client` kwarg. Both
kwargs are first-class in slack-sdk and slack-bolt; no new
dependencies. Resulting header looks like:
HermesAgent/<version> Python/3.x slackclient/3.x ...
No behavioral change for users — the Slack API ignores User-Agent
semantically; it lands in logs and analytics. Reversible.
Tests in tests/gateway/test_slack.py:
- TestSlackUserAgent pins the prefix shape and runs connect()
end-to-end (multi-token config) to assert every AsyncWebClient
carries the prefix and AsyncApp receives the pre-built client.
- TestSlackProxyBehavior fakes updated to tolerate the new kwargs
via **_kwargs so they don't break on future passthroughs.
.get("key", {}) only applies the default when the key is ABSENT.
When the key exists with value None (null in JSON), .get() returns
None and the subsequent .get() raises AttributeError.
Fix: replace .get("key", {}).get(...) with (.get("key") or {}).get(...)
which handles both missing keys AND None values.
8 instances across 6 files:
- gateway/run.py: tool_call function name check
- acp_adapter/server.py: tool name/description extraction
- gateway/platforms/qqbot/onboard.py: API response task_id
- gateway/platforms/yuanbao.py: message content parsing (x2)
- gateway/platforms/slack.py: block text extraction (x2)
- tui_gateway/server.py: error message extraction
Blank (or whitespace-only) answers to the home-channel prompt in the
interactive setup wizards previously left any previously saved
*_HOME_CHANNEL / *_HOME_ROOM env value in place, so operators could not
clear a stale home channel by re-running setup. Strip the prompt input
and call remove_env_value() on blank answers across the Discord, Slack,
Feishu, Matrix, Mattermost, WeCom and WhatsApp plugin setup wizards,
with per-adapter wizard tests covering set/clear/whitespace flows.
Squash of the three commits from PR #58421 (setup-wizard fix, matrix/
wecom extension, and 6-adapter test coverage) — one commit per
contributor on this salvage branch.
Fixes#12423
Salvaged from #58421
The C13 log-noise test (merged while this branch was in flight) builds a
bare SlackAdapter without __init__; the new _channel_teams ambiguity map
crashed it. getattr-guard, same pattern as the sibling defensive inits.
Five regression tests for _resolve_download_token and the download
helpers: explicit team wins, URL-embedded team id routes to the owning
workspace, unknown/no-match fall back to the primary token, and an
end-to-end _download_slack_file_bytes call asserts the Authorization
header carries the owning workspace's token.
A/B: all five fail with the fix commit reverted, pass with it applied.
Salvaged from #59742 (downloads half only). Main already resolves the
event-level team id from the channel→workspace cache before downloads
(the file-EVENT half was covered by #30456), but when neither the event
nor the cache knows the workspace, both download helpers silently fell
back to the PRIMARY workspace token — Slack then returns an HTML login
page instead of file bytes for any non-primary workspace file.
Slack private file URLs embed the owning workspace id
(files-pri/<TEAM_ID>-<FILE_ID>/...), so _resolve_download_token now
prefers: explicit team_id -> URL-embedded team id -> primary token.
Both _download_slack_file and _download_slack_file_bytes route through
it.
Deferred from #59742 (out of this correctness cluster's scope): the
channel→team disk persistence, the pre-send conversations.info probe
loop, and the thread-file ingestion feature.
Reapplied from #59742 by @benjamin2026-dot onto the current adapter
(original targeted the pre-#30456 download sites).
Conflict-resolution follow-up composing #68925 (Bob) with the already-
applied #20583/#66398 (jordanhubbard) recovery design:
- #68925's caller-level second _query_recoverable_session pass (via
lookup_session_key=) referenced a variable that no longer exists —
the legacy exact-key fallback now lives INSIDE
_query_recoverable_session, which also claims the legacy key once per
process and rewrites the peer row to the scoped key. Drop the dead
caller-level pass.
- Keep #68925's _recovered_row_matches_source_scope origin guard wired
into both recovery paths: a scoped channel lookup refuses rows whose
recorded origin names another workspace (or no workspace at all).
- Routing-index migration adoption policy documented at the site:
origin names a workspace -> exact match only; scope-less DM -> first
workspace claims once; scope-less channel -> refuse.
Two tests on main pinned the pre-#20583 bare-ts marker mechanism
(_mentioned_threads entries and _slash_command_contexts key shape).
The salvaged workspace scoping intentionally changes both:
- _mentioned_threads now records (team_id, ts) markers when the event
carries a team id, so the top-level-mention test asserts the scoped
tuple.
- _slash_command_contexts keys are (team_id, channel, user) 3-tuples
when the slash payload includes team_id.
Follow-up to the #20583 cherry-pick (jordanhubbard).
origin_session_id (the api_server wake self-post target) lived only in
the in-memory record: durable dispatch persistence and abandoned-
delegation recovery omitted it, leaving completions recovered after a
process restart unroutable to api_server sessions. Persist it in the
async_delegations table (CREATE TABLE + ALTER TABLE migration for legacy
DBs), restore it on recovery, and expose it via get_durable_delegation.
Also adds the contributors mapping for ianks (PR #64998 author).
Follow-up to #64998 (sweeper review F3).
/v1/runs bound only session_key at its _bind_api_server_session call, so
tools.async_delegation._current_origin_session_id() — which reads the
request-scoped HERMES_SESSION_CHAT_ID — returned "" on that route and
runs-originated background delegations stayed forced-sync with no wake
target. Bind chat_id/session_id the same way the other agent-entry routes
do via _run_agent(). Follow-up to #64998 (sweeper review F2).
On non-push adapters (api_server) the wake self-post IS the delivery, but
the cursor advanced before the self-post ran and a failed/exhausted post
was swallowed by the best-effort except — permanently losing the event.
Reorder the else-branch: for non-push adapters run the self-post FIRST and
only advance the cursor once it succeeds. A failure rewinds the pre-send
claim (same guarantee as the existing SendResult(success=False) path) so
the next tick retries, with the same MAX_SEND_FAILURES drop threshold.
Push-capable adapters keep the pre-existing advance-then-best-effort-wake
behavior. Follow-up to #64998 (sweeper review F1).
Wake-ups for kanban notifications and background delegation completions were
injected via handle_message() using a build_session_key()-derived key, which
can never match the raw X-Hermes-Session-Id key that api_server sessions run
under — so the wake landed in a session nobody was reading. On top of that,
ApiServerAdapter.send() reports failure without raising, and that was treated
as a successful delivery, so the notify cursor advanced past events that were
permanently lost; and background delegation was forced synchronous on
api_server since there was no way to wake the session afterward.
Fix: route wake-ups for non-push adapters through a self-post to
/v1/chat/completions with the original session id, treat non-raising send
failures as failures (rewind instead of advancing the cursor), and re-enable
background delegation whenever a session id is available to wake.
The origin session id is captured from the request-scoped api_server chat_id
binding rather than HERMES_SESSION_ID: constructing a child agent calls
set_current_session_id() with the subagent's internal id, clobbering that
variable right before dispatch would read it and misrouting the wake into
the subagent's own session.
Related: #56580, #64609, #53027, #63169, #56531, #50319, #64113
Correction to the previous commit (PR #68402): the claim that api_server
never intercepts MEDIA: tags is inaccurate on current main.
_resolve_media_to_data_urls() (gateway/platforms/api_server.py) DOES
inline image MEDIA: tags (<=5MB, image extensions only) as base64 data
URLs on the four main endpoints (_handle_session_chat,
_handle_session_chat_stream, _handle_chat_completions, _handle_responses).
The real gaps elphamale's PR points at are narrower:
- the /v1/runs output path (_handle_runs) never calls the resolver;
- non-image filetypes are never resolved anywhere (_MEDIA_IMG_EXT is
image-only).
Reword the hint to teach both halves: images via MEDIA: work on the
chat/completions/responses endpoints; non-image files and anything on
the runs endpoint must fall back to plain file paths in the response
text. Update the test to pin the scoped guidance instead of a blanket
prohibition.
Every PLATFORM_HINTS entry for a messaging platform (Telegram, WhatsApp,
Discord, Slack, Signal, WebUI, desktop) teaches the model the MEDIA:/path
convention because an interception mechanism actually resolves it there
(native attachment delivery, or a validated/inlined data URL). The cli
entry, which has no such mechanism, explicitly tells the model NOT to use
it and to state the path in plain text instead.
The api_server entry had neither instruction. Its /v1/runs handler never
routes the final response through any MEDIA: resolver (confirmed against
source: none of the four call sites of the api_server module's media-tag
resolver are inside its runs-endpoint handler), so a MEDIA:/path tag there
renders as inert literal text in the API response — exposing a raw host
filesystem path to the caller with no delivery ever taking place. Nothing
platform-specific told the model not to use a convention it's correctly
taught for several sibling platforms in this same dict, so the general
cross-platform habit could surface here too, unlike cli where an explicit
prohibition already closes the gap.
Mirrors cli's prohibition, adapted for api_server's actual constraint: no
"state the path in plain text" fallback, since a typical API caller has no
filesystem access to the host at all. Points at "a registered file-delivery
tool" generically rather than naming any specific tool, since api_server
toolsets are deployment-defined.
Follow-up to the salvaged PR #39439:
- Replace the hardcoded English gateway-auth summary with a
notifications.errors.gatewayAuthFailed i18n key (en/ja/zh/zh-hant + types)
- Fix the malformed ['"'] regex character classes (duplicate quote) in
notifications.ts error matchers
- Add regression test: gateway_auth_failed maps to the gateway auth
summary, provider invalid_api_key still maps to the OpenAI summary (#39365)
The api_server adapter returned error code "invalid_api_key" for
API_SERVER_KEY authentication failures, which the Desktop error
classifier misidentified as a provider (OpenRouter/OpenAI) key
problem — showing "OpenRouter API key missing" when the real issue
was gateway auth.
Changes:
- gateway/platforms/api_server.py: return "gateway_auth_failed" code
with descriptive message for API_SERVER_KEY auth failures
- apps/desktop/src/store/notifications.ts: add "gateway_auth_failed"
handler before "invalid_api_key" to show correct error message
- agent/error_classifier.py: add "gateway_auth_failed" to auth patterns
- tests: update test_session_api.py to expect new error code
Fixes#39365
The gateway ignored config when defined through
YAML (config.yaml). The API server only started when environment
variables like and were set,
even though other gateway subsections like were
processed correctly.
Two changes in gateway/config.py's load_gateway_config():
1. Merge platform configs placed directly under gateway.*
(e.g. gateway.api_server) via _merge_platform_map, matching
the existing gateway.platforms.* and platforms.* merge
paths. Only keys matching known Platform values are picked up;
non-platform keys like streaming are safely ignored.
2. Bridge api_server-specific keys (port, key, host, cors_origins,
model_name) from the top-level config block into the extra
dict so PlatformConfig.from_dict preserves them — matching what
_apply_env_overrides already does for env var values.
When _api_key_passes_startup_guard() rejects the key (missing,
placeholder/too short, or fail-closed unverifiable strength), connect()
returned a bare False with no fatal-error info. gateway.run's reconnect
watcher treats that as transient and re-queues with backoff forever —
each retry re-instantiating the adapter and its ResponseStore sqlite
connection. Observed in production (#37011): ~501 leaked connections
(1002 fds) over ~2.5 days until EMFILE made the whole gateway
unresponsive.
Set a non-retryable fatal error (api_server_key_invalid) in connect()
when the guard rejects, covering all three rejection branches, so the
platform drops from the reconnect queue; recover with
`/platform resume api_server` after fixing the key. Same treatment as
the port-conflict guard (api_server_port_in_use, #65665 / bda8bd76a8).
Tests mirror the port-conflict precedent: each rejection path asserts
connect() is False, has_fatal_error True, fatal_error_retryable False,
and fatal_error_code api_server_key_invalid, plus a strong-key control.
Re-implementation of #38803 by @cifangyiquan against current main —
their patch targeted the old inline guard in connect() which was since
extracted to _api_key_passes_startup_guard() (and gained the
fail-closed branch in 683059feb5), so the original diff no longer
applies. Their production diagnosis and non-retryable direction
preserved.
Refs: #38803, #37011
Salvaged from PR #36180 (commits 68dfeb4b16 and 86f437509b by arimu1),
re-applied onto current main with the incidental black-reformat churn
stripped out (~1,700 lines -> the semantic change + tests).
Previously gateway/config.py enrolled the api_server platform on
`api_server_enabled or api_server_key`, so API_SERVER_ENABLED=true with
no key (or a weak/placeholder key) still loaded the platform: the
adapter is instantiated (ResponseStore/SQLite opened in __init__), the
reconnect watcher spins, and the startup guard refuses at connect() —
logging errors forever. Now the platform is enrolled only when
API_SERVER_KEY passes the same strength bar as the adapter's startup
guard (has_usable_secret, min_length=16), via a shared
_has_usable_api_server_key() helper.
The no-op `lambda cfg: True` connected-checker for API_SERVER is also
replaced with the same key check, so get_connected_platforms() only
reports the platform "up" when it could actually start.
Known limitation (intentionally out of scope): a YAML config with
`platforms.api_server.enabled: true` and no key still loads the
platform; this gate covers the env-override path only.
Dropped from the original PR: EMAIL/SMS checker additions (scope creep
beyond the PR title; absent on current main) and the wholesale black
reformat of gateway/config.py and tests.
Fixes#36111
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
shivasymbl, z23, Skywind5487, 2001Y, gonzalofrancoceballos, kylezh.
briandevans' noreply mapping already present; mzkarami's mapping file
shipped inside #66204's own commit.
Also fixes the TestFormatMessageTableIntegration fixture to match
PlatformConfig's current signature (no 'name' kwarg).
Widen #48476's fence guarantees to the two splitters that still emitted
fence-broken chunks:
* GatewayStreamConsumer._split_text_chunks (fallback final send): close
the orphaned ``` at each chunk boundary and reopen it — with the
original language tag — on the next chunk, mirroring
BasePlatformAdapter.truncate_message's contract. Headroom is reserved
so balanced chunks stay within the platform limit.
* Slack block_kit._split_text (3000-char section chunking): same
close/reopen balancing for mrkdwn section text carrying fences.
With these, every chunk boundary — non-streaming send
(truncate_message), streaming overflow (_truncate_for_stream via
adapter.truncate_message per #45938), fallback final
(_split_text_chunks), final-send balance (ensure_closed_code_fences),
and Block Kit section splits — delivers fence-balanced chunks.
Regression tests probe each path with fenced fixtures, assert per-chunk
balance, limit compliance, language-tag reopening, and prose passthrough.
Slack's Block Kit `markdown` block accepts standard markdown (tables,
headers, task lists, fenced code with syntax highlighting, links) and
lets Slack translate it natively — eliminating the lossy markdown→mrkdwn
conversion for the rendered layout. Enable via
platforms.slack.extra.markdown_blocks.
Safety rails added on top of the original design:
* opt-in (default off) — Slack documents the block for 'apps that use
platform AI features' and does not guarantee availability across all
app types / surfaces, so unconditional adoption is not safe yet
* the mrkdwn-converted text field is ALWAYS kept as the
notification/search/accessibility fallback
* content over Slack's 12k cumulative markdown-block cap declines to the
rich_blocks renderer / plain text path
* the existing block-rejection retry (invalid_blocks / msg_too_long /
too_many_blocks) re-sends the plain mrkdwn payload, so an unsupported
surface degrades gracefully instead of dropping the message
* when both modes are enabled, markdown_blocks is preferred over the
local rich_blocks renderer; rich_blocks remains the fallback
Adapted from #8554 by @shivasymbl — the original patched the deleted
gateway/platforms/slack.py and switched unconditionally; reimplemented
against the plugin adapter's _maybe_blocks/sanitize_blocks pipeline.
Fixes#8552.
Slack mrkdwn has no table syntax — GFM pipe tables render as literal-pipe
noise with a raw |---|---| separator row. Wrap detected tables in ```
fences so they render as monospace preformatted text, and pad cells to
per-column max display width (East-Asian Wide / Full-width chars counted
as 2 columns) so columns stay aligned even with CJK content.
Tables already inside fenced code blocks are left untouched, and the
emitted fences carry no language tag so they compose with the
lang-tag-strip pass. This covers the plain-mrkdwn text path; the opt-in
rich_blocks path already renders native Block Kit table blocks.
Reapplied from #16648 by @kylezh — the original patched
gateway/platforms/slack.py, which was migrated to
plugins/platforms/slack/adapter.py in the plugin migration.
Fixes the non-rich_blocks table path described in #8552.
ensure_closed_code_fences previously only handled triple-backtick
(```) code-block fences. Single backtick (`) inline-code spans have
the same problem: an orphaned opening backtick causes the remainder
of the message to render as inline code on Discord and other platforms.
After balancing triple-backtick fences, strip complete ```...```
regions and count remaining standalone backtick markers. If odd,
append a closing backtick. Same trade-off as the triple-backtick fix:
a stray closing backtick may create a brief empty inline-code span,
which is far less harmful than the rest of the message being inline code.
Adds `ensure_closed_code_fences()` helper to detect text with an odd
count of triple-backtick markers (indicating an unclosed code block)
and append a closing fence.
Applies the fix to all four identified gap paths:
G1: truncate_message early-break path for final chunk
G2: _send_or_edit streaming edit path (most commonly hit)
G3: overflow split first chunk (covered by G2's fix)
G4: _send_fallback_final fallback send path
Closes: #TBD
B1: When reasoning content contains ``` e.g. model quoting code in
its thinking, wrapping it in an outer ``` for display causes the
inner fence to break the outer block.
Adds escape_code_fences_for_display() in gateway/stream_consumer.py,
called from gateway/run.py before wrapping reasoning in the outer
``` display block.
Slack's mrkdwn parser can fail to recognize the closing * of a bold span
when it is immediately preceded by a non-word character (), ], }, ., :,
em-dash, ...), mis-rendering the span and in reported cases truncating
the rest of the message. Insert a zero-width space (U+200B) between the
last character and the closing * whenever the last character is not
alphanumeric or underscore.
Reapplied from #35144 by @gonzalofrancoceballos — the original patched
gateway/platforms/slack.py, which was migrated to
plugins/platforms/slack/adapter.py in the plugin migration.
Slack's mrkdwn does not strip the optional language tag from fenced
code blocks like GitHub-flavored markdown does — it renders
```text\nfoo\n``` as a code block whose literal first line is "text".
The agent emitted ```text fences around raw command output, which
surfaced "text" as the first line of every such block.
Drop the tag from the opening fence in format_message() before stashing
the block behind a placeholder. Stripping only fires for a genuine
opening fence — a ``` at the start of a line, tagged with a single
token (no spaces or backticks) — and the original line ending is
preserved. The fence-protection regex deliberately matches loosely, so
a mid-line ``` (e.g. an inline ```span``` wrapping across a newline)
can be grouped as an "opening fence" whose first line is real content;
differential fuzzing against the pre-change formatter (40k generated
messages) confirms the only behavioral delta is the tag strip itself.
The Block Kit renderer is unaffected: render_blocks() intercepts fences
itself before mrkdwn_fn is applied, so this only changes the mrkdwn
surfaces that still go through format_message() — the plain-text
fallback field (notifications, search indexing, accessibility),
slash-command ephemeral replies, and standalone cron delivery.
Originally written against gateway/platforms/slack.py; ported to
plugins/platforms/slack/adapter.py after the adapter migration in
5600105478.
Manually verified against a live Slack workspace (pre-migration
adapter; the ```text case strips identically) — code blocks no longer
carry a literal "text" first line.
format_message unescapes already-escaped input before re-escaping, so that
pre-escaped text doesn't get double-escaped. That unescape was three
sequential str.replace calls, which re-scan each other's output:
"&lt;" --(& -> &)--> "<" --(< -> <)--> "<"
The & produced by the first replace pairs with the following "lt;" and
decodes a second time. "&lt;" is the wire form of the literal text
"<", so the text is silently destroyed: Slack receives "<" and renders
"<". Anyone writing about HTML or markup ("&lt;b&gt;" -> "<b>")
loses their literal text, with no error.
re.sub scans left-to-right and never re-scans its own replacements, so a
single pass fixes it. The escape pass on the next line is left untouched --
it is correctly ordered (& first, so the &s it inserts aren't re-escaped).
Only the double-decode cases change; every other input is byte-identical
before and after. This is the same round-trip invariant the neighbouring
test_pre_escaped_{ampersand,lt,gt}_not_double_escaped tests already assert,
extended to the case they miss. Affects the plain mrkdwn path (send,
edit_message) and Block Kit sections, which route section text through
format_message.
Test doubles and third-party adapter objects don't all implement the new
set_reaction_handler; calling it unconditionally hung the multiplex
secondary-reconnect test (the AttributeError was swallowed into an
awaited-forever path). getattr-guard all three wiring sites — same
duck-typing convention the sibling setters rely on for MagicMock, but
explicit, so bare objects work too.
Build the full reaction pipeline on top of the #29916 base:
- Opt-in gate: slack.reaction_triggers (default OFF — reaction events
stay acked-and-dropped so busy channels don't wake the agent on every
emoji). 'true' routes reactions on the bot's OWN messages; an explicit
emoji-name list routes those emojis from any message (handoff flows).
- reaction_removed events now route too, distinguished by the
cross-platform text convention reaction:added:<emoji> /
reaction:removed:<emoji> (matches the Feishu and Photon adapters, so
agents and skills see one shape everywhere).
- Authorization: the reactor becomes the synthesized message's user, so
the early _is_user_authorized gate and allowed_channels whitelist
apply exactly as for typed messages. _hermes_force_process only skips
the mention requirement (a reaction on the bot's own message is
definitionally addressed to the bot), mirroring Feishu/Photon.
- Gateway hooks (#33111 by @johnkattenhorn): every human reaction on a
message item fires reaction:added / reaction:removed through the new
BasePlatformAdapter.set_reaction_handler → GatewayRunner
._handle_reaction_event → HookRegistry.emit, independent of the
routing opt-in. Documented in hooks.md.
- Channel handoff (#45265 by @Kev-fs): slack.reaction_trigger_target
routes the reaction turn to a configured channel (top-level via
_hermes_no_thread_response + reply-anchor suppression in
gateway/platforms/base.py) or C123:<ts> thread.
- Manifest: reaction_removed event subscription added alongside
reaction_added/reactions:read.
- Docs: slack.md Reaction Triggers section; hooks.md event table rows.
Also credits #44508 by @harrisonmedmedmetrics (inbound reaction_added
handling — same plumbing class, superseded by this consolidated shape).
Co-authored-by: johnkattenhorn <john.kattenhorn.personal@gmail.com>
Co-authored-by: Kev-fs <kevin@fleetsmarts.net>
Co-authored-by: harrisonmedmedmetrics <harrison@medmetricsrx.com>
Slack reaction_added events were explicitly acked and dropped, so a user
reacting to a bot message (👍 to approve, ✅ to acknowledge) produced
nothing. Forward them through the normal message pipeline as synthesized
MessageEvents whose text is the reaction emoji (translated to unicode
for common names), keeping the downstream auth gate, thread-context
fetch, dedup, and skill routing unchanged.
- Self-reactions and non-message items are dropped; reactions on
messages not sent by this bot are dropped (Feishu-adapter parity).
- The reacted-to message's thread parent becomes the synthesized
thread_ts so the reaction lands in the same session as a reply would.
- Manifest gains reactions:read scope + reaction_added bot event.
Salvaged from PR #29916 by @bpross.
Related: #33111, #44508, #45265 (same cluster).
Salvaged from PR #36220, ported onto the current SessionStore (SQLite-
backed get_or_create_session; activity check is last_prompt_tokens) and
the sidecar-note reset path (context notes now ride turn_sidecar_notes
instead of prepending to context_prompt).
Long-lived Slack/Discord channels/threads lose their context on
daily/idle session resets, and the agent can bind a new request to an
unrelated recent session (observed: a Discord thread reset caused a PR
in the wrong repository). Record prev_session_id when an auto-reset
replaces a session with real activity, persist it, and append a
deterministic one-line hint to the auto-reset context note pointing the
agent at session_search for that specific prior session. No LLM calls,
no channel-history APIs, no extra DB lookups; other platforms and
activity-free resets are untouched.
Refs #36220. Co-authored-by: metamon <269728612+metamon-p@users.noreply.github.com>
Salvaged from PR #45702 (heartbeat half). A multi-minute turn showed a
static 'is thinking...' assistant status that reads as stuck and provokes
mid-turn 'you there?' pings. Derive the fallback status label from the
turn's elapsed time (>=30s → 'still working… (NmSSs)'), riding the
existing _keep_typing refresh — zero extra API calls.
Ported onto the current plugin adapter: the start time rides the tracked
_active_status_threads entry (workspace-scoped key), so it shares the
existing bounds/eviction and resets when stop_typing clears the status.
Explicit live-status phrases (set_status_text) and configured
typing_status_text always win; only the built-in default label changes.
The PR's other half (top-level channel follow-up coalescing) is NOT
included — dispatch semantics changed on main (busy-input active-turn
redirect, #30170 demotion) and need a fresh design pass.
Refs #45702. Co-authored-by: MrAbsaroka <mrabsaroka@gmail.com>
Issue #47699 reported Slack /goal continuations being enqueued by
_post_turn_goal_continuation -> _enqueue_fifo but never drained until the
next real inbound message woke the session. On the current tree the
continuation lands in the adapter pending slot while the
_process_message_background frame is still live, so the in-band pending
drain (and the finally-block late-arrival drain) spawns the follow-up
turn automatically — the reported stall is not reproducible on main.
Pin the two halves of that contract so it can't silently regress:
1. a continuation placed in the FIFO during the handler frame is
consumed as a second turn without any new user message, and
2. the runner's goal hook enqueues under the same session key the
adapter drain resolves (key mismatch would orphan the event).
Refs #47699. Reported-by: joesu-angible
process(wait) marks a completion consumed and returns the exit code +
output inline. The gateway process watcher's agent-notify branch honored
that (skipping the synthetic agent turn), but its skip FELL THROUGH to
the plain text-notification branch, which re-sent the same completion to
the chat as a raw '[Background process ... finished with exit code ...]'
message — a duplicate delivery of output the agent had already read and
was summarizing (observed on Slack with
display.background_process_notifications: all, but platform-agnostic).
Guard the raw-notification branch on is_completion_consumed(), same as
the agent-notify branch. poll() stays read-only and never marks consumed
(#10156), so status checks still can't suppress autonomous delivery.
Fixes#65379. Reported-by: hergert
* fix(desktop): hide persisted agent-only history scaffolding
Filter verification-stop nudges and context-compaction handoffs at the
stored-history mapper boundary. Preserve a real reply when a compaction
handoff shares its stored message.
* test(desktop): build persisted E2E sessions through the real agent
Drive tui_gateway.entry over its stdio JSON-RPC transport against the mock
provider, wait for real completion events, and persist normal session history
through AIAgent and SessionDB. Migrate resume and hidden-history coverage,
including real compression and live verify-on-stop scaffolding, then remove
the unused direct SessionDB import scripts.
* fix(desktop): use the provisioned Python for real-session E2Es
Run the stdio gateway through uv's synced project environment outside the
Nix dev shell, while retaining the fully provisioned Nix Python when the
shell advertises HERMES_PYTHON_SRC_ROOT.
* fix(nix): expose the provisioned Python environment to uv
Mark the Nix-built Python environment active in the dev shell so the shared
E2E session builder can always run through `uv run --active --no-sync`.
* fix(timeline): persist typed display events
* fix(timeline): strip display-only fields from provider payloads, preserve through rewrites, fix /resume display history
Three review findings from PR #69771:
1. Provider payload leak: display_kind and display_metadata were forwarded
to the provider API as unknown message fields. Strict OpenAI-compatible
backends can reject the next request after a model switch or resumed
typed event. Strip both from the per-request api_msg copy in
conversation_loop alongside the existing api_content pop.
2. Rewrite/import data loss: _insert_message_rows preserved display_kind
but silently dropped display_metadata. After replace_messages,
archive_and_compact, or session import, async-delegation completion
events lost their task counts and fell back to generic display text.
Add display_metadata to the INSERT columns and bind tuple.
3. CLI /resume stale recap: startup --resume A set _resume_display_history
from A's lineage. A subsequent in-session /resume B loaded B only into
conversation_history via get_messages_as_conversation, leaving the stale
A display projection. _display_resumed_history preferentially read the
stale attribute, showing A's recap for B. Switch /resume to
get_resume_conversations and update _resume_display_history alongside
conversation_history.
Tests: 890 Python (5 files), 35 desktop TS — all green.
* feat(tui): render typed display events as ◈ markers in the Ink TUI
The TUI was not handling display_kind at all — model switch markers and
async delegation completions rendered as opaque user messages with the
full [System: ...] text, and hidden compaction handoffs were visible.
Wire display_kind through the full TUI chain:
- _history_to_messages (tui_gateway/server.py) forwards display_kind
and display_metadata to the gateway transcript payload.
- GatewayTranscriptMessage (gatewayTypes.ts) gains both fields.
- Msg.kind (types.ts) gains 'event' value.
- toTranscriptMessages (domain/messages.ts) maps:
- hidden → skip entirely
- model_switch → event "model changed"
- async_delegation_complete → event "N background agents finished"
(or "background agent work finished" without metadata)
- messageGroup (blockLayout.ts) routes event to its own group, with
SELF_SPACED + PAINTS_TRAILING_GAP so it owns its margins.
- messageLine.tsx renders event-kind as a dim ◈ marker with no gutter,
matching the CLI's ◈ event rendering.
- 4 new TUI tests for hidden/model_switch/async_delegation mapping.
TUI typecheck: clean. TUI lint: 0 errors (2 pre-existing warnings).
TUI tests: 9 passed (1 pre-existing failure on main, unrelated).
Verified line-by-line against agent/bedrock_adapter.py on this branch:
allowlist split (Nova primary, Claude bearer-token fallback only,
everything else rejected), exact system→tools→messages[-2] insertion
order, and the usage normalization formula/field mapping.
Claude-on-Bedrock already gets prompt caching via the AnthropicBedrock
SDK path. This adds it to the raw Converse API path used for non-Claude
models (Amazon Nova, and Claude when bearer-token auth forces Converse
routing, #28156) — a conservative model allowlist inserts cachePoint
blocks after tools, system, and the message before the newest turn, and
extracts cacheReadInputTokens/cacheWriteInputTokens into usage so cost
accounting picks them up through the existing Anthropic-style fallback.
Ref: relatorio-cache-performance-provedores-ia.md, P0 item 1.
Follow-up hardening on top of the C14 cherry-picks (#57860/#44026/#66742/#60009):
- Slack file downloads (_download_slack_file/_download_slack_file_bytes)
now require an https URL on a Slack CDN host (files.slack.com,
*.slack.com Enterprise Grid, *.slack-files.com legacy shares) before
attaching the bot token. url_private/url_private_download only ever
point at the Slack CDN, so a forged file object from a malicious
workspace app or compromised event stream pointing the Bearer-token
download at an arbitrary PUBLIC host (token exfiltration) is now
refused — a hole #44026's generic private-IP SSRF check alone could
not close.
- The same two download paths now use create_ssrf_safe_async_client
(from #57860) so the preflight-validated hostname is resolved once,
validated, and dialed by IP — closing the DNS-rebinding TOCTOU window
for the token-bearing inbound fetches as well.
- #60009's slack_tokens.json permission warning is generalized into
utils.warn_if_credential_file_broadly_readable() (POSIX-only,
fail-quiet) and wired into the other read path with the same gap:
google_chat's load_user_credentials(). google_chat already writes
0o600 via _write_private_json; the read-time warning covers
hand-provisioned/legacy files. Nothing in-repo writes
slack_tokens.json (user/OAuth-provisioned), so there is no write
path to chmod for Slack.
Security tests both directions: non-CDN/lookalike/http URLs and
connect-time DNS rebinds are blocked before any TCP connect; real
files.slack.com, Enterprise Grid, and slack-files.com URLs still reach
the network layer; 0o600 files stay silent while 0o644/0o640 warn with
a chmod hint. A/B: all 10 new download-guard tests fail with the
hardening reverted and pass with it applied.
The OAuth multi-workspace token file contains plaintext bot tokens for
all saved Slack workspaces. Unlike the Google Chat adapter which sets
0o600 when writing credentials, the Slack token file has no permission
enforcement — a default umask 022 makes it world-readable.
Fix: check file permissions on read and emit a warning log with remediation
instructions if the file is group- or world-readable.
`SlackAdapter._fetch_thread_context` formats each prior thread message as
`{name}: {msg_text}` and joins them with newlines into the block the call
site prepends *raw* into the model turn (`text = thread_context + text`).
Both fields are attacker-influenceable — any thread participant sets their
own Slack display name and message text — and neither was neutralized, so
an embedded newline let a thread message break out of its line and pose as
a fresh markdown section (a fake "## SYSTEM" / "## Override" heading) inside
the context the model reads when first mentioned mid-thread.
This is the same indirect-prompt-injection vector already closed for the
sibling untrusted sinks: the sender-name prefix
(`neutralize_untrusted_inline_text`), the reply quote, and the relay
channel-context renderer. The Slack thread-context backfill — the default
whenever the bot is mentioned in a thread with no active session — was the
missed sink. (The existing `[unverified]` tagging marks *who* a message is
from; it does nothing about newline structure, so an authorized sender can
inject just as easily.)
Flatten both fields with `neutralize_untrusted_inline_text` before
interpolation. The body uses `max_chars=0` so message text is not truncated
(thread context caps the message *count*, never per-message length); the
display name keeps the default bound. `parent_text` keeps the raw message
(its own reply-context sink neutralizes separately). A well-behaved message
is preserved byte-for-byte.
Adds a regression test covering a hostile display name, a hostile message
body, benign passthrough, and the no-truncation guarantee.
_download_slack_file and _download_slack_file_bytes fetched Slack-supplied URLs with the bot token attached and follow_redirects=True, but without the is_safe_url pre-flight check or per-redirect guard that the outbound send_image path already uses. A URL that resolves to (or 3xx-redirects into) a private/internal address could reach internal services and leak the bot token (CWE-918). Add the same pre-flight + _ssrf_redirect_guard hook to both inbound sibling paths.
Several platform fetch paths called is_safe_url before constructing ordinary httpx clients, leaving a second DNS lookup at connection time. This preserved the rebinding window for Slack batch images, Feishu documents, Telegram URL-photo fallback, and WeCom remote media.
Route each path through create_ssrf_safe_async_client and the shared redirect guard so direct connections validate and dial vetted IPs while configured proxies remain an explicit trusted egress boundary. Add per-path regressions that change DNS from public at preflight to metadata at connect time.
The Skills Hub provenance fixture intentionally serves content over loopback. Opt that test-scoped server into private-address access so it keeps exercising the real HTTP transport without weakening production blocking.
Related #8033
Co-authored-by: teknium1 <127238744+teknium1@users.noreply.github.com>
Install connect-time DNS validation for Hermes-owned direct httpx clients so SSRF-sensitive fetch paths dial a vetted IP instead of re-resolving after preflight. This preserves Host/SNI semantics for direct HTTP(S) connections and keeps proxy routing as an explicit trusted egress boundary.
Wire the guarded clients into media cache downloads, vision downloads, Skills Hub direct/raw fetches, and platform attachment fetch paths that already perform SSRF preflight and redirect validation.
Fixes#8033
Co-authored-by: Tom Qiao <zqiao@microsoft.com>
CI shards run without the slack extras; the two #64267 tests import the
real SDK (SlackApiError, the lazy-rebind path) and errored with
ModuleNotFoundError. Skip on bare environments — classification coverage
for stdlib exception types (OSError/TimeoutError/cert errors) still runs
everywhere.
Widening for #24117 ('is thinking...' stuck after the response was
sent): the stuck-thinking class is a missing-cleanup-on-error-path bug.
send()'s status clear was gated on thread_ts already being resolved and
on reaching the normal post-message path, which left the Slack
Assistant status visible when a turn ended through any sibling exit:
- exception BEFORE _resolve_thread_ts (slash-context handling,
formatting, DM resolution) — the 'if thread_ts: stop_typing' clear
never ran
- empty/whitespace-only final response (no_text guard early-return)
- ephemeral slash replies (Slack only auto-clears assistant status on
real thread replies; ephemerals never count)
Add _clear_thread_status_quietly() — a best-effort stop_typing wrapper
that never masks the caller's SendResult — and wire it at every send()
exit plus the finalize paths of edit_message. stop_typing already
handles the untracked-thread fallback (clearing an unset status is a
no-op on Slack's side), so this is pure coverage widening.
Also add the #18859 unit tests for _resolve_progress_thread_id's new
reply_in_thread gate (synthetic-thread drop, real-thread keep, event-id
fallback suppression, default unchanged).
A/B: 3 of the 4 new status-clear tests fail with the adapter widening
reverted and pass with it applied; the fourth pins the existing
cleanup-must-not-mask-result contract.
The Slack adapter honours platforms.slack.extra.reply_in_thread=false
in _resolve_thread_ts, but the Gateway's progress-message path forced
event_message_id as the thread_id for Slack regardless. The first
progress message ('terminal: …', 'Processing…') created a thread that
all subsequent edits and the final answer inherited, defeating the
user's reply_in_thread=false setting.
Check the live Slack adapter's reply_in_thread flag before applying the
event_message_id fallback, and treat a synthetic source.thread_id (==
the event's own message ts, used only for session keying) as 'no
thread' so progress messages stay at the channel/DM top level.
Folds both #18859 commits (reply_in_thread gate + synthetic thread_id
drop) into main's extracted _resolve_progress_thread_id helper — the
original patched the pre-refactor inline block; the gate now composes
as a keyword argument so Mattermost/other platforms keep the default
fallback behavior.
When reply_in_thread=false, top-level channel events carry their own
message ts as metadata.thread_id for session keying. Calling
assistant.threads.setStatus on that ts activated a Slack assistant
thread ('is thinking...') before the actual response was sent, and the
flat reply then never cleared it.
send_typing now routes through the same _resolve_thread_ts synthetic-
thread guard as message sending, and the gateway threads message_id
through progress/status metadata so the adapter can distinguish real
threads from synthetic top-level session keys.
Reapplied from #18859-sibling PR #17184 by @dorukardahan (both commits:
fix + progress-metadata test) onto current main via 3-way apply — the
original patched gateway/platforms/slack.py, moved to
plugins/platforms/slack/adapter.py in the plugin migration.
Progress/status callbacks (context-pressure, compression retries,
model fallback) route through _send_or_update_status_coro, which
edits the previous bubble for the same status_key when the adapter
implements send_or_update_status — but only Telegram did. On Slack
every status event posted a fresh thread message, so a compression
retry loop spammed a dozen out-of-order bubbles into the thread
('Context too large 1/3... 2/3... 3/3', fallback switches, etc.).
Implement send_or_update_status on the Slack adapter following the
Telegram pattern (#30045): first call posts and caches the message ts
per (channel, thread, status_key); subsequent calls edit that message
via chat.update. Edit failure drops the cached ts and falls back to a
fresh send. Cache is FIFO-bounded.
Covers cluster C1-images (#69185, #32315, #66136):
- _slack_file_marker unit tests: typed markers per mimetype family,
hostile-filename sanitization (newlines/brackets can't fake context
structure).
- _render_message_text appends markers; a caption-less image post no
longer vanishes from thread context.
- Cold-start hydrate integration: prior-message images surface as
markers in channel_context; the thread root's image is downloaded,
delivered as media_urls, and upgrades message_type to PHOTO.
- Failure path: root-image download failure degrades to the marker,
never blocks the turn.
- Bounds: root delivery capped at _THREAD_ROOT_IMAGE_MAX; non-image
root attachments stay marker-only (no download).
- One-time delivery: active thread session skips the hydrate → no
re-download/re-delivery on later turns.
- Composition: the trigger's own event files still ride alongside a
delivered root image; Slack Connect stubs resolve via files.info;
the collector never issues its own conversations.replies call.
- Delta refresh (#23918 path): images in new replies past the watermark
surface as markers, with no root re-download.
A/B: 14 of 15 tests fail with the adapter fix reverted, all pass with
it applied.
When the bot is mentioned mid-thread for the first time, the thread root
is very often the artifact the mention is about ("@bot what's in this
chart?" posted as a reply under an image) — but the root's image never
reached the agent, so it answered blind.
On the cold-start hydrate path (and only there), _collect_thread_root_images
reads the root message from the thread-context cache the immediately
preceding _fetch_thread_context call just populated (zero extra Slack API
calls in the normal case), downloads its image/* attachments through the
existing authenticated _download_slack_file helper, and delivers them as
media_urls/media_types on the same MessageEvent — upgrading the message
type to PHOTO so vision routing engages.
Scope and safety:
- One-time delivery by construction: the cold-start path is guarded by
_has_active_session_for_thread, so later turns in the same session can
never re-download or re-deliver. No new gateway/session plumbing needed.
- Bounded by _THREAD_ROOT_IMAGE_MAX (4); non-image root attachments stay
text-only markers.
- Slack Connect stubs (file_access=check_file_info) resolve via files.info.
- Best-effort: a failed download degrades to the [image: ...] marker from
the thread context — never an error turn.
- Also hardens the video mimetype fallback (mimetype can be empty) so
media_types entries are always non-None strings.
Adapted from #69185 by @KCAYAAI — the original plumbed MessageEvent media
through gateway/base.py, run.py and session.py with durable one-time
delivery markers (2,441 lines); this lands the user-visible behavior
adapter-locally by reusing the session guard already on the hydrate path.
Images and files posted in a Slack thread before the bot joins were
invisible to the agent: _fetch_thread_context renders text only, and a
caption-less image post was dropped from context entirely (empty text →
skip). "@bot what do you think of the chart above?" read as a question
about nothing.
_render_message_text now appends a compact, sanitized marker per file
attachment — [image: chart.png], [video: demo.mp4], [audio: note.m4a],
[file: report.pdf (application/pdf)] — so the agent can SEE that prior
thread messages carried attachments and ask for a re-share when it needs
the bytes. Filenames are stripped of newlines/brackets so a hostile name
can't fake context structure. Because both thread-context formatting and
parent-text rendering go through _render_message_text, markers appear on
the cold-start hydrate, the explicit-mention delta refresh, restart
rehydration, and reply_to_text.
Reapplied from #32315 onto the current adapter (original patched the
pre-plugin gateway/platforms/slack.py, moved in the plugin migration;
annotation labels reworked to per-file typed markers, download side
handled separately).
#58478's caplog test predates main's unlabeled-bot users.info probe
(#69xxx wave-2 gating): events without client_msg_id now hit
_resolve_user_is_bot, which the fixture's mock client doesn't wire up
(AttributeError on _user_is_bot_cache). Real human-authored Slack
messages carry client_msg_id — add it to the fixture so the test
exercises the intended block-extraction path.
Follow-up hardening for the C13 log-noise cluster:
- plugins/platforms/slack/adapter.py: clarify button resolution logged
the full chosen option text at INFO (choice=%r). Choice text is user
content — log the choice INDEX at INFO and the (truncated, %.100r)
text at DEBUG only. Widens #58478's principle: no message content
above DEBUG level anywhere in the adapter.
- tests/gateway/test_slack_log_noise.py (new): behavioral suite pinning
the cluster's invariants:
* catch-all ack registered AFTER every named handler (registration
order is bolt's dispatch priority — no shadowing);
* catch-all fires for an unsubscribed event type (reaction_added),
logs only a DEBUG line naming the type, and never logs content;
* named handlers still dispatch (message → _handle_slack_message);
* end-to-end inbound message run leaves NO message text or block
content in any adapter log record (caplog at DEBUG);
* #30185's event-arrival diagnostic is metadata-only;
* clarify resolution: INFO carries index/user only, text is
DEBUG-only (fails with the adapter fix reverted — A/B verified).
Content-leak audit of all 139 logger call sites in the adapter found
two above-DEBUG leaks: the clarify choice line (fixed here) and none
else carrying message text; remaining sites log error strings, URLs
via safe_url_for_log, ids, and counts. The block-extraction DEBUG
preview was already removed by #58478 (chars= length only).
Regression test for the catch-all ack: a re.compile(r'.*') event matcher
must be registered (after every named handler, so it never shadows
message/app_mention/reaction/file routing) and must match unhandled
subscribed event types like member_joined_channel / channel_archive /
pin_added.
Salvaged from #64218 (test half only — its adapter-side catch-all is a
duplicate of #38847, which landed as the base commit of this cluster
with first-submitter credit). Fixes#6572.
Co-authored-by: shivasymbl <sdevinarayanan@asymbl.com>
Slack posts are durable workspace messages, not an ephemeral terminal
status area. Default long_running_notifications and busy_ack_detail to
off for Slack so long-running agent work does not leave permanent
operational breadcrumbs like 'Working — 9 min — iteration 12/90' in
channels. Both remain opt-in per platform via
display.platforms.slack.*.
Also covers the platform-generic shutdown-notification mute path with a
regression test (gateway_restart_notification=false must suppress both
the active-session interruption notice and the home-channel copy).
Salvaged from #69028 (quiet-defaults half only). The PR's other half —
the channel_session_scope_channels session-scoping feature — is a new
config feature outside this log-noise cluster and overlaps the session
scoping territory reworked by merged wave-1/2 Slack session work; it is
deliberately not taken here.
Without a catch-all handler, slack-bolt returns HTTP 404 for every
unhandled bot event (user_change, user_huddle_changed, reaction_added,
etc.) and never sends the Socket Mode ack. On active Slack workspaces
where the app is subscribed to high-volume events, this produces a
near-100% un-acked failure rate that crosses Slack's >95%/60-min
threshold and triggers automatic disabling of the app's Event
Subscriptions — silently killing all inbound event delivery.
Place a catch-all re.compile(r".*") handler AFTER the specific event
handlers so bolt's router matches those first. Truly unhandled events
are silently acked (200) and logged at DEBUG. The failure rate stays
near 0% regardless of which events the Slack app manifest subscribes to.
Fixes#6572
Treat Slack users.conversations missing_scope as an expected limited-scope app condition and fall back to session history without recurring warnings.
Add tests for not-ok and SlackApiError-like missing_scope responses.
Follow-up to the #68627 cherry-pick (cluster C15 — Slack platform
capability-note accuracy; earliest report/fix: #6545 by @daikeren):
1. Session/prompt stability: the pinned session-context render
(_pinned_session_context_prompt) is keyed by _ephemeral_change_key,
whose contract requires every rendered input to appear in the key.
The new _slack_tools_loaded() gate reads config + the live MCP
registration map, so its state is now hashed into the key exactly
like the existing Discord gate — a gate flip re-renders ONCE (a
legitimate bust); within a session the note stays byte-stable for
the life of the conversation (A/B: the new parity test fails with
this key change reverted, passes with it).
2. Derived, non-overpromising positive note: rather than hardcoding a
capability list that can drift stale again (the original bug class),
the tools-present note tells the agent to consult the actual loaded
Slack tool schemas for supported operations — the schemas ARE the
source of truth, so the note cannot overclaim ops a given Slack
toolset/MCP server doesn't expose (e.g. a read-only history server).
3. Tests: parity test proving a gate flip changes both render and key;
byte-stability test proving three consecutive turns in one Slack
session return the identical pinned object (sha256-equal); autouse
fixture pins the new gate so key<->render parity is env-independent.
Ports #63234 forward onto current main per teknium1's review.
gateway/session.py hard-coded the stale-API disclaimer for every Slack
session regardless of whether Slack tools were actually loaded. This
contradicted the system prompt when MCP or native slack tools were
present, causing the agent to refuse Slack API actions it could
actually perform (issue #6536).
Per review, the original predicate only checked the native 'slack'
toolset, missing Slack MCP servers (registered under mcp-<server> in
tools/mcp_tool.py) entirely. _slack_tools_loaded() now checks two
independent paths:
1. Native 'slack' toolset + SLACK_BOT_TOKEN (as before, but now calls
_get_platform_tools() with include_default_mcp_servers=True instead
of False, so a default-enabled MCP server also counts).
2. A connected MCP server that has ACTUALLY registered tools into the
live registry (new tools.mcp_tool.get_registered_mcp_server_names()),
whose name suggests Slack. This is session-scoped in the sense that
matters here: MCP servers connect once per gateway process (not
per-session), so checking the live per-server tool-registration map
is the correct availability-filtered signal -- unlike the earlier
get_all_tool_names() approach this replaces, which conflated ALL
built-in tool names process-wide, this only inspects the small,
purpose-built MCP server-name map.
Added a real regression test that registers a tool via the actual
tools.mcp_tool._track_mcp_tool_server() tracking function (not a mock
of the capability check) to verify a genuine Slack MCP server is
detected, plus a negative case for an unrelated MCP server.
5/5 Slack-specific tests pass; 126/126 in the full
tests/gateway/test_session.py file.
Follow-up to the salvaged PR #57320 commit: the PR bridged port/host/secret
for MSGRAPH_WEBHOOK but only tested webhook and api_server. Adds a test
covering the msgraph_webhook branch including extra-precedence, plus the
contributors/emails mapping for kjames2001.
WebhookAdapter and ApiServerAdapter read port/host from config.extra, but
PlatformConfig.from_dict only populates extra from the 'extra:' sub-key in
the YAML platform section. Top-level keys like port and host are silently
ignored, causing the adapter to fall back to DEFAULT_PORT (8644).
This causes silent port conflicts in multi-profile setups: a profile that
configures 'platforms.webhook.port: 8649' still binds 8644, colliding with
the default profile's webhook on the same port.
Fix: extend the shared-key bridging loop in load_gateway_config() to bridge
top-level port/host/secret into extra for WEBHOOK, MSGRAPH_WEBHOOK, and
API_SERVER platforms, following the same pattern already used for
dm_policy, allow_from, gateway_restart_notification, and other keys.
The extra dict takes precedence: if port is already under 'extra:', the
top-level value does not clobber it.
Documents user-facing wave-1+2 Slack behavior that had no docs coverage:
- decision table for require_mention / free_response_channels /
require_mention_channels / thread_require_mention / strict_mention /
ignore_other_user_mentions and how they compose
- 'Accepting messages from other bots' section with the post-#69483
semantics: allow_bots=mentions requires a CURRENT mention from
peer bots (text or Block Kit blocks); thread state never admits them
- clarify one-tap buttons (choice buttons + Other free-text mode,
in-place resolution, double-click guard, expiry message)
- slash replies are ephemeral: replace-ack, chunking, 5-post cap with
explicit truncation notice, postEphemeral fallback, never-public rule
- cron deliver targeting (slack -> home channel, slack:C... channel,
slack:U... resolved to DM) incl. standalone sender + MEDIA uploads
- send_message media + bare-user-ID DM resolution and caption behavior
Refs #26184.
Salvaged from #45765 by @navahc09 — kept the PR's callout structure
and placement, rewrote the content to match current behavior:
Slack blocks native slash commands in threads and never delivers them,
so Hermes recognises a leading '!' as an alternate command prefix.
Post-C3 command fixes the bang form also works behind a mention
(@Hermes !cmd) and with leading whitespace; unknown '!' tokens pass
through to the agent unchanged. Cross-linked the detailed
slack.md section.
0.15s missed by 1-8ms on busy CI shards twice today (runs 30025889952,
30026770278 — branches not touching this file). The assertion pins
'handler did not block on the timeout path' (blocking = seconds); 2.0s
keeps the contract per the loose-bounds flake policy.
A user who configured a provider only inside a MoA preset (advisor or
aggregator slot) has explicitly opted into that provider — the consent
gate (is_provider_explicitly_configured) now scans moa.reference_models,
moa.aggregator, and all moa.presets.* slots, so Claude Code OAuth pool
seeding and the auxiliary auto-fallback chain treat MoA-only Anthropic
users consistently with model.provider users.
Salvaged from PR #57778 (trimmed): the auxiliary_client fallback half of
the original PR was independently landed on main in ddd3a2d247 and is
dropped here; a secret-scrubber artifact in the gate test fixture is
restored to the real placeholder token.
_apply_env_overrides() force-set ``api_server.enabled = True`` whenever
API_SERVER_KEY (or API_SERVER_ENABLED) was present in the environment.
In multiplex mode, a secondary profile pins
``platforms.api_server.enabled: false`` in its config.yaml so that it
shares the default profile's API-server listener instead of binding its
own port. That profile still inherits the process-level env, including
API_SERVER_KEY, so the unconditional re-enable flipped api_server back on
and tripped the MultiplexConfigError check.
Honor an explicit disable, flagged by ``_enabled_explicit`` in the
platform's extra. Use ``extra.pop("_enabled_explicit", False)``: the
api_server branch is terminal (unlike the migrated plugin platforms, no
later registry pass re-enables api_server), so popping consumes the flag
in a single read and avoids the double-read hazard, while the final
per-platform cleanup remains a no-op.
Adds a regression test asserting that with API_SERVER_KEY set, a config
with api_server explicitly enabled:false + _enabled_explicit:true survives
_apply_env_overrides() as enabled=False (fails without the fix), while the
key is still wired through for the shared listener.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Follow-up to #60168's salvage: aggregate_moa_context() is the third
independent MoA call path; assert its aggregator call receives the
custom-provider request_overrides.extra_body via **agg_runtime.
'auto' is a sentinel meaning "inherit from main runtime / auto-detect",
not a literal model id -- already handled for cfg_model (config-derived)
in _resolve_task_provider_model, but not for the explicit `model` kwarg.
MoA reference/aggregator slots (agent/moa_loop.py's _slot_runtime) forward
a preset's `model:` field as this explicit argument rather than through
auxiliary.<task> config, so a MoA preset configured with `model: auto`
(a natural thing to try given the existing auxiliary.*.model: auto
convention) reached this function as the explicit `model` arg and took
the `model or cfg_model` branch, bypassing the cfg_model-only sentinel
check entirely -- sending the literal string "auto" to the wire as a
model id.
Normalize both the explicit `model` and `cfg_model` the same way, fixing
this at the single chokepoint every caller (MoA included) already goes
through, rather than patching moa_loop.py separately.
Extends the desktop backend's root-cause fix (aa2ae36c3f) to all remaining
console-less parent launch paths. The Windows console-flash class
(#54220/#56747) is governed by the PARENT's console: a DETACHED_PROCESS or
pythonw.exe daemon has no console, so every console-subsystem descendant
(git, gh, cmd, node, wmic, powershell) allocates its own visible conhost —
one flash per spawn, unreachable by any per-call-site CREATE_NO_WINDOW
sweep. Worse, MSDN specifies CREATE_NO_WINDOW is IGNORED when combined
with DETACHED_PROCESS, so the hide bit in the old detach bundle was dead.
Changes:
- _subprocess_compat: drop DETACHED_PROCESS from windows_detach_flags()
and windows_detach_flags_without_breakaway(); the daemon now owns a
single hidden console (CREATE_NO_WINDOW) that all descendants inherit.
- gateway_windows: _resolve_detached_python() returns the venv console
python.exe (no pythonw/base-interpreter detour — the uv-shim flash
premise only held while DETACHED_PROCESS was masking the hide bit);
UAC handoff launches console python under SW_HIDE; cmd/vbs launchers
render console python (vbs runs it window-style 0).
- gateway/run.py: restart watcher keeps sys.executable instead of
swapping in GUI-subsystem pythonw.
- web_server: dashboard actions spawn sys.executable (already carries
windows_detach_flags()).
Tests updated to pin the new invariants, including an explicit
DETACHED_PROCESS-must-stay-out regression guard.
Follow-up to the #62614 salvage: try_refresh_matching (added by the
#69843 salvage after this PR's base) calls self.current() while already
holding the now-locking non-reentrant pool lock — a guaranteed deadlock
that git merges silently (no textual conflict). Use _current_unlocked()
and cover the method in the no-deadlock test.
Follow-up to review feedback:
- Acquire self._lock in the remaining public pool-state methods:
has_credentials, reset_statuses, remove_index, resolve_target, and
add_entry. All of them read or rebind self._entries (and the mutating
ones persist auth.json), so they now hold the same lock as select()
and the query methods. None are called from within the lock, so no
unlocked helpers are needed.
- Make the blocking test deterministic: an instrumented lock records the
acquire attempt, and the test first waits for the worker to actually
reach self._lock before asserting it blocks. Previously an unlocked
method could pass if the worker thread was scheduled late.
- Extend the lock test matrix to all nine public methods; the five newly
locked ones fail the test without this fix.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
`has_available()`, `peek()`, `current()` and `entries()` read (and, via
`_available_entries()`, mutate and persist) `self._entries` without holding
`self._lock`, while every other entry point — `select()`,
`mark_exhausted_and_rotate()`, `acquire_lease()`, `try_refresh_current()` —
guards the exact same access with the lock.
`_available_entries()` is not read-only: it prunes aged-out DEAD manual
entries (rebinding `self._entries` at the prune step) and calls `_persist()`
(writes auth.json). The gateway runs platform adapters in threads and cron
runs jobs in a ThreadPoolExecutor, so a status probe via `has_available()`
or `peek()` can race a concurrent `select()`/rotation: torn iteration of
`self._entries`, interleaved auth.json writes, or a lost token rotation.
Fix: take `self._lock` in all four query methods. Because the lock is
non-reentrant and `peek()` composes `current()` + `_available_entries()`,
add a lock-free `_current_unlocked()` helper and route the already-locked
internal callers (`_select_unlocked`, `mark_exhausted_and_rotate`,
`_try_refresh_current_unlocked`) through it to avoid self-deadlock.
Added regression tests: a no-deadlock check (peek re-entrancy) and a
lock-held-blocks-the-call check for each of the four methods.
Decompose children inherit the root's literal workspace_path (#37172),
so every sibling of a worktree-kind root points at the SAME checkout.
_resolve_worktree_workspace's existing-checkout shortcut then reuses
that directory on whatever branch is currently checked out, ignoring
the task's own branch_name. Net effect: sibling workers — which can be
promoted and dispatched concurrently — run in one directory on the
first sibling's branch, with no lock. Work lands on the wrong task's
branch (provenance corruption) and concurrent siblings trample each
other's index/tree.
Fix, two layers:
- decompose_triage_task: worktree-kind children no longer inherit the
root's literal path; each child materializes its own
<repo>/.worktrees/<child-id> at dispatch (dir/scratch inheritance
unchanged — children legitimately share those).
- _resolve_worktree_workspace: when the requested path is an existing
checkout of a DIFFERENT branch, fall back to a fresh
<repo>/.worktrees/<task-id> instead of silently reusing it (heals
rows that already carry a shared path). Same-branch reuse and the
no-repo/own-path degenerate cases keep the legacy behaviour.
Tests: tests/hermes_cli/test_kanban_worktree_isolation.py (5); full
test_kanban_db.py + test_kanban_decompose_db.py suites pass unchanged.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Follow-up to the #65844 salvage: the new anthropic pool test must stub
read_claude_code_credentials like the sibling tests, otherwise a dev
machine's live claude_code singleton seeds a third entry and the
no-benching assertion fails outside CI.
Two related races in credential-pool cooldown state:
1. Lost update across processes: write_credential_pool merged only
entries missing from the caller's snapshot; for entries present on
both sides the caller's in-memory copy won wholesale. A process
holding a snapshot taken before another process marked a key
exhausted would, on its next persist (e.g. a round-robin rotation),
write the key back as healthy — erasing the cooldown so every
process resumes hammering a rate-limited key. Merge status fields by
last_status_at recency: adopt the on-disk status only when it is
strictly newer AND still binding (DEAD, or EXHAUSTED with an
unexpired cooldown), and never onto re-authed (token-changed)
entries, so legitimate expiry-clears and fresh logins are preserved.
2. Wrong-key quarantine: when mark_exhausted_and_rotate received an
api_key_hint that matched no entry, it fell through to
current()/_select_unlocked() — on a freshly loaded pool that selects
the NEXT healthy key and benches it for the full cooldown TTL,
punishing an innocent credential. When a hint is provided but
unmatched, rotate without marking anything instead of guessing.
Includes regression tests.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
git stash push --include-untracked exits non-zero when it saved
everything but could not DELETE some swept untracked files from the
working tree (e.g. a root-owned packaging/ directory left behind by a
sudo'd build: 'warning: failed to remove ...: Permission denied').
The updater ran the push with check=True, so this benign partial
failure raised CalledProcessError and aborted the whole update before
it even fetched — reliably, on every run, for any user with an
undeletable untracked path in the checkout.
Fix, both ends of the class:
- _stash_local_changes_if_needed: probe refs/stash before/after the
push. Non-zero push + fresh stash entry = changes are saved; warn,
reset the tracked-side leftovers (they're in the stash), and
continue the update. Non-zero push + NO stash entry = real failure;
keep aborting.
- _restore_stashed_changes: on restore, those same undeletable files
still sit in the tree, so 'git stash apply' exits 1 with 'already
exists, no checkout' even though every tracked change applied and
nothing was lost. Classify that stderr shape (strictly — any other
error line still routes to the conflict path) as restored instead
of resetting the tree and telling the user the restore failed.
Repro'd both halves with real git; behavioral E2E test covers
stash -> checkout -> restore round-trip with an undeletable dir.
A 402/429/401 is an API-key–level failure (account out of balance,
rate-limited, or key rejected), but the same key can back more than one
pool entry — e.g. an explicit pool entry plus a `model_config` entry
auto-seeded from `model.api_key`, both carrying the identical
`runtime_api_key`.
`mark_exhausted_and_rotate(api_key_hint=...)` only marked the *first*
matching entry, leaving the sibling OK. `_select_unlocked()` then kept
handing back the same depleted key, so the billing-recovery `continue`
loop in the conversation retry path never converged: the request hung
until the client disconnected (~2.5min observed against DeepSeek),
emitting only `response.created` with no 402 ever surfaced to the user.
Mark every entry sharing the failed key so the pool can reach the
"no available entries" state and let the error propagate immediately.
Adds a regression test covering two entries backed by the same key.
The dedup-reset calls assumed a full AIAgent; gateway/loop test doubles
built via object.__new__ lack _clear_context_overflow_warn and crashed
in build_turn_context (caught by test_api_content_sidecar on CI slice 3).
getattr-guard all four call sites per the established test-double pitfall
pattern (AGENTS.md #17).
Follow-up fixes for the #62625 salvage:
- Dedup-reset gap (sweeper review): when the block clears while the
context is STILL over threshold, execution enters the compression
branch — the PR's 'else' reset never ran, so the warning stayed
suppressed forever after the first block. _clear_context_overflow_warn()
now fires on every automatic compression path: turn-context preflight,
conversation_loop pre-API gate, and the post-tool loop-compaction gate.
- should_compress_info on current main: main refactored should_compress
into _automatic_compression_blocked()/_locally(); the tuple variant now
derives its reason from the same in-memory state via
_compression_block_reason(), keeping cooldown:<s>/ineffective shapes.
- ContextEngine.should_compress_info ABC default now actually returns
(should_compress(tokens), None) — the PR's default had a docstring but
no return (returned None, would crash tuple-unpacking call sites).
- Below-threshold guard: the turn-context persisted-cooldown branch and
the conversation_loop pre-API cooldown branch no longer warn when the
estimate is under threshold (should_compress_info returns a None
reason; the preflight pre-check is not a threshold guarantee). The
pre-API guard also honors compression.max_attempts instead of a
hardcoded 3, and no longer fabricates a cooldown reason.
- Noise-filter survival (#69550 composition): warning text is now a
template constant (CONTEXT_OVERFLOW_BLOCKED_WARNING_TEMPLATE) marked
FAILURE-CLASS, pinned un-swallowed in VISIBLE_COMPRESSION_MESSAGES and
in new tests that execute the real _TELEGRAM_NOISY_STATUS_RE +
_prepare_gateway_status_message.
- Contributor mapping for stanislav@local -> sl4m3.
- ContextEngine.should_compress_info() default impl so plugin engines
(e.g. _StubEngine) don't raise AttributeError at the call site.
- Centralise warning/reset in AIAgent._warn_context_overflow_blocked /
_clear_context_overflow_warn so turn-context and conversation-loop guards
share identical dedup logic and reset on the real compression boundary.
- Cover conversation_loop.py pre-API (~L1007) and loop-compaction (~L4774)
guards, not just the turn-context preflight.
- _FakeAgent mirrors the two helpers; test suite green (219 passed).
Fixes#62708
Previously, when a session crossed the compression threshold but compression
was skipped (summary-LLM cooldown, #11529, or anti-thrashing, #40803), the
model kept accumulating context until it hit the hard provider token limit and
silently stopped answering — with no signal to the user about why.
Changes:
- context_compressor.should_compress_info() returns a (should_compress, reason)
tuple. reason is 'cooldown:<seconds>' or 'ineffective' when compression is
needed but blocked. should_compress() keeps its bool contract so existing
callers (conversation_loop.py) and regression #29335 are unaffected.
- turn_context.build_turn_context() emits a deduped _emit_warning when the
context is over threshold but compression is blocked, advising /new or
/compress. Dedup keys on the block *kind* (cooldown/ineffective), not the
ticking countdown, so a cooldown doesn't re-fire the warning every turn.
- Adds tests/agent/test_turn_context_overflow_warning.py covering the tuple
shape, both block kinds, dedup, and re-fire-after-clear.
Default kanban_create children now keep fresh scratch paths, while explicit dir sharing remains supported and project context resolves to a per-task worktree. Surface resolved workspace fields in create responses/events and cover scratch mutation, nesting, explicit sharing, and project inheritance.
Fixes#67567
Dispatcher-spawned Kanban workers are finite one-shot processes, so detached delegation completions can outlive their only consumer. Mark that runtime as unable to deliver async completions and reuse the synchronous delegation fallback, returning required child results before the worker exits.\n\nAlso make unsupported-session notes runtime-generic and cover the delayed-child lifecycle regression.\n\nRefs #63169
Extend the existing candidate-name resolver in _supports_vision_override
to accept 'vision' as an alias for 'supports_vision' on per-model config,
for both the providers.<name>.models dict and the legacy list-style
custom_providers form.
Per review feedback on #31912: this extends the current resolver rather
than replacing its candidate-name logic. Named custom providers resolve
to the runtime value provider='custom' while the config keeps the
user-declared name under model.provider; that lookup path is preserved.
Adds regression tests covering model.provider=my-vllm with runtime
provider='custom' for both config shapes.
The bare truthiness test on _compression_skipped_due_to_lock is fooled
by MagicMock auto-attributes on test-double agents (skill pitfall:
MagicMock defeats hasattr/truthiness duck-typing) — the type-ahead CLI
test's MagicMock agent took the lock-skip branch and skipped the
transcript commit. Real values are None/True/holder-string; pin the
check to 'is True or isinstance(str)' at all three consumer sites.
Follow-up to the salvaged #57634 commits:
- agent/manual_compression_feedback.py: new describe_compression_lock_skip()
— single source of truth for lock-skip wording. A descriptive holder
string means another compressor CONFIRMED holds the lock ('already in
progress (holder: ...)'); True/None means acquisition failed without a
confirmed holder (hermes_state.try_acquire_compression_lock catches
sqlite3.Error internally and returns False), so the message says
'could not acquire ... the lock check failed' instead of falsely
claiming a concurrent compression is running.
- cli.py, gateway/slash_commands.py, tui_gateway/server.py (all three
in-process consumers: session.compress RPC, command.dispatch compress
branch, slash.exec mirror) now route through the shared helper.
- tui_gateway/server.py command.dispatch compress branch: catch
CompressionLockHeld explicitly — it previously fell into the generic
'compress failed' error handler.
- Deferred-notify contract (#69324): lock-skip discards the pending
context-engine notification (committed=False) in _compress_session_history
and the CLI path before returning.
- tests: lock-skip wording pins per surface, VISIBLE_COMPRESSION_MESSAGES
noise-filter carve-outs for both wordings, MagicMock signal opt-outs for
sibling tests added on main after the original PR.
Advisor review found a critical stale-signal leak: if auto-compress
sets _compression_skipped_due_to_lock during a lock-skip, a subsequent
successful manual /compress will see the stale signal, falsely report
'Compression already in progress', and discard the compression results.
Fix:
- compress_context clears _compression_skipped_due_to_lock = None at
entry so each call's outcome alone determines the signal.
- Unified gateway 'holder: unknown' drift to match CLI/TUI pattern
(omit holder clause when not a descriptive string).
- Added MagicMock opt-outs in 3 sibling test files broken by the new
signal check (test_compress_here, test_compress_focus,
test_compress_plugin_engine).
- Added stale-signal-leak invariant test proving the fix.
Treat cua-driver's Linux `is_on_screen: null` as unknown instead of
off-screen, and skip GNOME Shell desktop/backdrop helper windows
(ding "Desktop Icons", @!x,y;BDHF) when selecting the default capture
target — they are targetable X11 windows but capture as empty.
Reconciled with the _NET_ACTIVE_WINDOW fallback from #58030: helper
windows are filtered out of the candidate pool first, then the tied
z-order active-window probe runs on the remaining real app windows.
Also falls back to the requested app name for _last_app when Linux
windows carry no app_name.
Salvaged from #54173 by @dnth.
The anti-thrash guard (_ineffective_compression_count) was in-memory
only: a fresh compressor bound to a resumed, already-compacted session
started with compression_count=0 and a disarmed guard, so a
near-threshold session could legally re-compact once per process
restart, forever.
Persist the counter through the durable session-state channel,
mirroring the failure-cooldown (#54465) and fallback-streak (af7dceaf7)
pattern:
- hermes_state.py: sessions.compression_ineffective_count column
(declarative reconciliation adds it on existing DBs) +
get/set_compression_ineffective_count accessors.
- context_compressor.py: every strike/clear verdict routes through
_record_ineffective_compression_verdict() which writes through to the
session row (no-change verdicts skip the DB write);
bind_session_state() loads the persisted value; the compression
rotation boundary carries the counter onto the child row;
update_model()'s reset also clears the durable copy; the
ineffective-only fast path in _automatic_compression_blocked() is
removed because the counter is now durable and another agent's clear
must unblock a stale local snapshot.
- conversation_compression.py: _refresh_persisted_compression_guards
re-reads the counter alongside cooldown + fallback streak.
Reset semantics are unchanged: any real provider reading below the
threshold still clears the counter — and now clears it durably too.
Resolves the residual gap identified in #54923 by @lanyusea (the
second-threshold mechanism was superseded by persisting the existing
guard state).
Co-authored-by: lanyusea <lanyusea@gmail.com>
The skill claimed Chromium/Electron 'reject synthetic clicks' so computer-use
can't drive the canvas. The Cua team disproved this on the exact v1.11.0
AppImage (Linux/X11): background delivery returns background_unavailable, but
that's the first rung, not a wall — cua-driver returns escalation:'foreground'
and its X11 XTest path (x11_xtest_fg) with delivery_mode:'foreground' clicks
through, dismissing the consent dialog and landing canvas clicks.
Corrected the note to say: climb to foreground on background_unavailable, don't
conclude Electron is unclickable. Ref: NousResearch/hermes-agent#67052.
Optional skill for driving the tldraw offline desktop app via its local
HTTP control API (the same curl-based path the app's own agent skills use
for Codex/Claude Code/Cursor/Gemini) — read the canvas, make live edits,
and write embedded document scripts.
Grounded in the app's bundled script-context.d.ts and agent playbook, and
in the real tldraw SDK v5 shape schema:
- document-script contract: export default function ({ editor, helpers, signal })
- HTTP API: /api/search, /api/doc/:id/exec, /api/doc/:id/script-workspace,
/api/doc/:id/script-status (bearer token from server.json, re-read per call)
- shape schema table validated against @tldraw/tlschema (scripts/validate_shapes.mjs, 3/3)
- interactive-UI example (scripts/counter.js) + diagram-generation (scripts/main.js)
- honest verification boundary: click->state logic verified via /exec dispatch
(0->1->2->1->0), with documented host caveats (inotify watcher, Electron
background-click rejection)
Tests: tests/skills/test_tldraw_offline_skill.py (15 passing).
`_api_key_passes_startup_guard` refuses to start the API server on a weak
`API_SERVER_KEY`, and its own log says why:
This endpoint dispatches terminal-capable agent work — a guessable key
is remote code execution.
But the check is wrapped so that a failure to import it starts the server
anyway:
try:
from hermes_cli.auth import has_usable_secret
if not has_usable_secret(self._api_key, min_length=16):
... return False
except ImportError:
pass
return True
`hermes_cli.auth` imports httpx at module scope and pulls in a large slice of
the CLI, so an import failure is not hypothetical — a trimmed image, a partial
install, or a circular import during gateway startup all produce one. When it
happens the strength check silently disappears and only the presence check
above it remains, so a placeholder key passes.
Reproduced against the real guard with the import blocked:
weak key, normal : False
weak key, ImportError : True <-- starts on a 4-char key
strong key, normal : True
Fail closed instead: an unverifiable key does not get to expose the endpoint,
and the log names the actual problem so the operator can repair the install.
This is the posture tools/credential_files.py already takes — it refuses a
mount when its deny-list cannot be consulted rather than risking it. The catch
also widens from ImportError to Exception, so an AttributeError or an error
raised inside the check cannot reopen the same hole.
Both happy paths are untouched: a strong key still starts, a weak or missing
key is still refused with the existing messages.
Unrelated to #38803, which fixes the retry behaviour after this guard rejects
and assumes the guard ran.
tests/gateway/test_api_server.py: new TestApiKeyStartupGuardFailsClosed — a
weak key is refused when the check is unavailable, a strong key is refused too
(fail-closed), plus three controls pinning the unchanged normal paths. The two
fail-open tests fail on main; the three controls pass there. 222 passed in the
api_server suites; 1475 passed across every suite touching api_server, with
the same 8 pre-existing failures on clean main.
Follow-up to srojk34's explicit-provider unwrap (PR #56691):
- Extract _resolve_moa_aggregator() as the single preset->aggregator
resolver shared by _resolve_auto(), _resolve_task_provider_model(),
and resolve_provider_client() so preset lookup/validation can't drift.
- When the main provider is moa, the aggregator model is now the default
for every UNSET auxiliary model: _read_main_model_for_aux() substitutes
the preset's acting (aggregator) model wherever fallback chains
pre-filled from _read_main_model() (router prefill, custom-endpoint
fallback, named-custom default, external-process default,
_try_main_agent_model_fallback).
- Unwrap moa at the resolve_provider_client() chokepoint so direct
callers (vision auto-detect, plugin code) can't dead-end in the
unknown-provider branch, and unwrap the vision auto-detect main
provider before capability probes run against the preset name.
- Real-config tests: temp HERMES_HOME + actual config.yaml exercising
the genuine load_config()/resolve_moa_preset() boundary.
_resolve_task_provider_model() returned an explicit provider="moa" override
(from a caller-passed arg, or auxiliary.<task>.provider: moa in config.yaml)
verbatim, with no MoA-preset unwrap. Only the *implicit* "main provider is
moa" path inside _resolve_auto() unwraps to the aggregator slot (#53827) —
this function never goes through _resolve_auto() at all, so the explicit
case was never covered.
MoA is a virtual provider with no real HTTP endpoint: resolve_provider_client()
looks "moa" up in PROVIDER_REGISTRY (no such entry), falls to the
unknown-provider dead end, and call_llm surfaces a nonsensical "Provider
'moa' is set in config.yaml but no API key was found. Set the MOA_API_KEY
environment variable..." error for a provider that was never meant to be
reached over the wire.
Fix mirrors #53827's aggregator-resolution approach exactly: when either the
explicit `provider` arg or the config-derived `cfg_provider` is "moa",
resolve the named (or default) MoA preset via resolve_moa_preset() and
continue with its aggregator's real provider+model, dropping any explicit
base_url/api_key (the moa:// virtual endpoint and placeholder key belong to
the facade, not the aggregator's real provider). If the preset can't be
resolved (renamed/deleted), degrades gracefully to the pre-fix behavior
instead of raising harder.
- agent/auxiliary_client.py: _unwrap_moa_provider() helper + call sites for
both the explicit-arg and config-derived provider="moa" cases in
_resolve_task_provider_model(). Also tightened base_url/api_key parameter
types to Optional[str] (matching their actual None-accepting behavior),
which incidentally resolved 5 pre-existing ty diagnostics at call sites.
- 5 new regression tests in tests/agent/test_auxiliary_client.py: explicit
arg unwrap, config-derived unwrap, default-preset fallback when no model
is configured, graceful degradation on preset-resolution failure, and a
non-moa regression guard.
The three subprocess tests spawn 'sys.executable -c' children that import
hermes_cli. From a worktree, the child resolved the MAIN checkout's editable
install instead of the tree under test, so the new DB/CLI guards appeared
missing and the tests failed with rc=0. Route the spawns through a helper
that pins the repo root under test on PYTHONPATH.
Mocked-subprocess tests asserting creationflags == CREATE_NO_WINDOW for
each path salvaged from PR #56877: tui_gateway cli.exec / shell.exec /
quick-command dispatch, the CLI quick-command exec handler, and the
Copilot ACP + Codex app-server Popen transports (pipes asserted intact).
Verified: all 6 fail with the fix reverted, pass with it applied.
Six spawn sites reachable from the desktop GUI / TUI gateway lacked CREATE_NO_WINDOW, so a windowless parent (pythonw/Electron) flashed a conhost per spawn: cli.exec RPC, quick-commands exec dispatch, and shell.exec RPC in tui_gateway/server.py; the CLI REPL quick-commands exec in cli.py; and the per-session provider transports in agent/copilot_acp_client.py and agent/transports/codex_app_server.py (Popen, hide-only so PIPE stdio stays intact).
All use hermes_cli._subprocess_compat.windows_hide_flags() (no-op on POSIX), matching the pattern already used at three other sites in tui_gateway/server.py. Deliberately hide-only — no detach flags, no Electron changes (per the #54220 revert history).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Follow-up to the #58738 salvage: the pre-exhausted check now enumerates
pool.entries() to find the failing key, so the MagicMock pool double must
expose entries as a callable, not a bare list.
Review follow-up: the auth path called pool.try_refresh_current() before
the hinted rotation, so a stale current() pointer could force-refresh a
different, healthy entry — consuming its single-use refresh token, or
(for non-OAuth entries, where a forced refresh marks the entry exhausted
outright) killing it entirely before api_key_hint was ever consulted.
Use try_refresh_matching(api_key_hint=...) to resolve and refresh the
entry that supplied the failing key under the pool lock, falling back to
the previous behavior when no key is known.
Adds a regression test with current() deliberately pointed at the
healthy entry: on the old code the healthy entry is exhausted by the
forced refresh and the pool ends up fully offline; with the fix the
failing entry is exhausted and recovery rotates to the healthy one.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
recover_with_credential_pool identified "which credential failed" via
pool.current(), a shared mutable pointer that is advanced by every
select() (round-robin rotation, concurrent turns, and other processes
reloading the pool reset it to None). By the time recovery ran, it
routinely pointed at a different, healthy entry — mark_exhausted_and_rotate
then stamped the failing request's error message and reset time onto that
innocent entry. With round_robin and one hard-capped key this
deterministically exhausted the healthy key too and took the entire pool
offline ("no available entries") from a single rate-limited credential.
mark_exhausted_and_rotate already supports api_key_hint for exactly this
(the auxiliary-client path passes it); the main conversation-loop path
never did. Pass agent.api_key — kept in sync with the entry in use by
_swap_credential — as the hint on all four rotation call sites, and make
the "already exhausted → rotate immediately" pre-check look up the failing
entry by key with the same fallback to current().
Adds regression tests that fail on the old attribution logic: a fresh
pool (current() is None) failing on key B must mark entry B, never
entry A.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The hermes console entry point is hermes_cli.main:main, and main.py imports
dotenv (via env_loader) and yaml (via config) at module level. In the #57828
failure state — a failed lazy backend refresh wiping a core package's import
files while metadata survives — a normal launch crashed while importing
main.py, before _recover_from_interrupted_install() and the recovery markers
from PR #58004 could act.
- hermes_cli/_early_recovery.py: stdlib-only bootstrap repair invoked at the
very top of main.py, before any third-party import. Probes the fragile
core packages via real imports, force-reinstalls broken ones using the
pyproject.toml pins, shares main.py's single-flight recovery lock, and
never clears markers (the confirmed lifecycle stays with the full recovery
path in main.py).
- Probe/repair tables now have one canonical home in _early_recovery, reused
by main.py so the two layers cannot drift.
- Manual --force-reinstall fallback commands now print pinned specs via
_lazy_refresh_repair_specs() instead of bare package names.
- tests: entry-point lifecycle coverage proving a broken dotenv import
crashes main.py without repair and imports cleanly with it, a stdlib-only
import guard for _early_recovery, and unit coverage for marker gating,
lock single-flight, pinned specs, and marker preservation.
Keep .update-incomplete for full .[all] recovery only. Lazy refresh uses
.lazy-refresh-incomplete and clears only after confirmed import probes;
unavailable probes are indeterminate, not healthy (#58004 review).
Keep .update-incomplete across normal hermes.exe launches, heal via
package-only import probes first, and only clear the marker after repair
succeeds (#57828 / #58004 review).
Upgrade pip before lazy refreshes, probe core imports when a lazy
install fails, force-reinstall corrupted packages with pyproject pins,
use package-only install (no shim quarantine) for repair, and keep the
.update-incomplete marker until refresh/repair succeeds (#57828).
The startup pruner only considered directories named hermes-* (the
hermes -w scratch trees), so salvage/review/port lanes created with raw
'git worktree add' accumulated forever — a real checkout reached 117
directories / 26 GB with trees dating back months. Two further leaks:
squash-merged branches' local commits stay unreachable from
refs/remotes/* forever, so the unpushed-commits guard preserved fully
merged scratch trees indefinitely; and preserved trees rotted silently
with no visibility.
- Pruner now covers every directory under .worktrees/ except kanban
task trees (t_<hex>, owned by 'hermes kanban gc'). Named (non
hermes-*) trees get a 3x timeline (72h soft / 9d hard) since they
were created deliberately.
- New _worktree_commits_all_merged_upstream(): git-cherry
patch-equivalence check against origin/HEAD|main|master, bounded at
20 commits ahead, fails safe toward preserve. Lets the pruner reap
trees whose every local-only commit already landed upstream via
squash-merge/cherry-pick.
- Dirty guard now applies at every tier (previously the 24-72h tier
skipped it — it only survived because the unpushed check usually
caught the same trees).
- Trees preserved for unpushed/dirty reasons older than 7 days are
listed in a single WARNING so in-flight work can't rot silently.
- tips.py text updated; 13 new behavior-contract tests.
- gateway/run.py: use _adapter_for_source(source) instead of the raw
adapters.get(source.platform) map so the compression-timeout warning
respects transport provenance, relay ingress, and multiplexed profiles
(matches every other user-facing send in the hygiene block).
- tests: add a lock-release verification regression — a fence-cancelled
hygiene compression must leave the per-session compression lock free so
the next attempt (manual /compress retry) acquires it and commits
normally.
Follow-up for the salvaged #35191: the mid-turn pre-API pressure emit in
conversation_loop.py and the idle-resume emit in turn_context.py were not
routed through automatic_compaction_status_message, so an engine with
emit_automatic_compaction_status=False still leaked those lines. Both now
resolve through the hook (phases "pre_api" and "idle") while keeping the
#69550 template constants as the default wording. Suppression also skips the
#69546 structured 'compacted' terminal edge for compress-phase events that
opened no visible phase; failure warnings (_emit_warning) remain never
suppressible, pinned by test.
Covers the follow-up hardening: continuation-marker and summary-as-user
tails keep the flagged standalone snapshot (zero-user provenance #69292
verified via _transcript_has_real_user_turn on the projected rows),
stale snapshot rows are refreshed in place, a previously merged snapshot
is stripped before re-injection, and an all-completed todo store injects
nothing (#26981).
Follow-up hardening on the salvaged merge-into-trailing-turn fix:
- Merge only into REAL user tails (_is_real_user_message probe). Merging
into scaffolding tails (continuation marker, summary-as-user handoff)
would upgrade them to real-user evidence after SessionDB projection
strips the flags, breaking zero-user provenance (#69292 -
_is_synthetic_compression_user_turn keys on the TODO_INJECTION_HEADER
content marker, which merge-at-tail would bury mid-content).
- Strip a previously merged snapshot block before re-injection so
repeated boundaries refresh rather than accumulate todo state, and
refresh a bare stale snapshot row in place instead of stacking a
duplicate (empty/stale-skip semantics from #26981 by @YLChen-007).
- Scaffolding tails keep the flagged standalone append (pre-#53890
status quo; adjacent user rows are repaired downstream by
repair_message_sequence / _merge_consecutive_roles).
After context compression, the preserved todo list was unconditionally
appended as a standalone user message. When the compressed transcript
already ends with a user message (common case), this creates consecutive
user/user turns — a role-alternation violation some providers reject.
Fix: fold the snapshot into the trailing user message (blank-line separated)
when one exists with plain-string content. Falls back to append when the
tail is non-user, empty, or has structured (list) content.
Rebased on current upstream/main.
Closes#53890
Regression test for the exact issue #61932 report: head + an 8-message
protected tail made exclusively of oversized tool pairs. Pre-fix,
compress_start >= compress_end made compress() a pure no-op and the
retry loop ended in 'Cannot compress further'; post-fix the Phase-1
pressure demotion reclaims the tail in one pass while preserving
tool_call/tool_result pairing.
After multiple in-place compactions, short tool-heavy sessions can leave
nearly every remaining message inside protect_last_n while those messages
are huge completed file/tool outputs. The middle compress window then
makes no material token progress and the turn dies with
"Cannot compress further" (#61932).
Cap the prune message floor at the same bound as tail-cut, and under
pressure demote bulky protected-tail tool bodies (keeping a short recent
floor) so preflight can reclaim headroom without wiping the active ask.
Follow-up for salvaged #58629: thread the previous flush baseline through
the idle-compaction caller in turn_context.py (the one caller the original
PR predates), and add regression coverage that every early-return path in
compress_context (breaker skip, no-progress, plus the completed-rotation
boundary) resets the per-attempt in-place outcome so a stale
_last_compaction_in_place from an earlier successful in-place compaction
can never baseline unflushed turns as persisted.
Relocates the #20424 wiring: the preflight region moved out of
run_agent.py into agent/turn_context.py (and through the
compression.max_attempts unification, #69315), so the contributor's
elif branch is reapplied at its current home as the else arm of the
threshold dispatch chain.
Integration contracts:
- Byte-identical default: the built-in ContextCompressor inherits
ContextEngine.should_compress_preflight() -> False, so the default
path performs no compression and touches no turn bookkeeping
(pinned by test_builtin_compressor_default_sub_threshold_path_unchanged).
- Attempt-cap: the engine gets exactly ONE compress() pass per turn,
mutually exclusive with the cap-bounded threshold multi-pass loop,
so turn-start passes stay within the resolved
compression.max_attempts budget in every case.
- No-op blocking (#64382 / 377244f7c): an engine pass that no-ops
(_compress_context returns the input list object) neither sets nor
clears preflight_compression_blocked and does not re-baseline the
flush history — a sub-threshold maintenance no-op proves nothing
about over-threshold compressibility.
- Engine exceptions are swallowed at debug level; cooldown/defer/
codex-native gates run before the hook is ever consulted.
Salvaged from #20424 by @Beandon13. Fixes#20316.
Context engines that override ``should_compress_preflight()`` (e.g. the
hermes-lcm plugin's incremental leaf-chunk compaction) never had their
hook fired by ``run_conversation`` because the preflight block exited
early once the hardcoded ``>= threshold_tokens`` check failed. As a
result, ``LCM_DEFERRED_MAINTENANCE_ENABLED=1`` and friends were inert
and accumulated raw_backlog debt indefinitely.
Add an ``elif`` branch that delegates to the engine's preflight hook
when the legacy threshold check does not fire. The default
``ContextEngine.should_compress_preflight()`` returns ``False`` so the
built-in ``ContextCompressor`` is unaffected; engines opting in get a
chance to ingest messages and request a single ``compress()`` pass for
deferred maintenance. Exceptions are swallowed at debug level so a
buggy engine cannot break an otherwise-healthy turn.
Closes#20316
- tests/cli/test_compress_type_ahead.py: end-to-end proof of the PR #68284
docstring claim — a prompt queued into _pending_input while /compress runs
survives compaction untouched and is the next item process_loop drains,
i.e. it is processed against the compacted history. Plus a structural
guard that handle_enter never gates on _command_running /
_command_blocks_input (read-only enforcement belongs solely to the
TextArea Condition; a busy-gate in handle_enter would silently drop
type-ahead input).
- tests/test_cli_manual_compress.py: update the one remaining _busy_command
stub (added on main after the PR branched) to accept the new
blocks_input kwarg.
- contributors/emails: map lucas@policastromd.com -> enzo2.
Follow-up to the salvaged #35533 fix: instead of parsing 'here [N]' only in
_mirror_slash_side_effects, parse it inside _compress_session_history — the
single helper all three manual-compress routes converge on (session.compress
RPC, command.dispatch /compress|/compact, slash-exec mirror). Every route now
honors the boundary-aware forms (here [N], up to here, --keep N) with the
same head/tail split + rejoin as cli.py and gateway/slash_commands.py, and
keeps the choke point's existing guards (lock-free LLM call, history_version
race check, deferred context-engine notification).
Tests: choke-point unit tests for 'here N', degenerate-split fallback, and
focus-topic passthrough, plus endpoint-level tests on each of the three
routes.
_mirror_slash_side_effects() passed the raw argument after /compress
directly as focus_topic to _compress_session_history. So /compress here
3 silently used "here 3" as a summary focus topic and did a full
compress instead of preserving the last 3 exchanges verbatim.
Fix: call parse_partial_compress_args() on the argument first. When it
detects a boundary-aware form (here, here N, up to here, --keep N),
split the history into head/tail using split_history_for_partial_compress,
compress only the head via agent._compress_context, and rejoin with
rejoin_compressed_head_and_tail — exactly mirroring cli.py and
gateway/run.py's /compress here implementation (PR #35252).
Non-boundary forms (plain /compress, /compress <focus>) fall through
to _compress_session_history unchanged.
Photon (managed iMessage) shipped without a _PLATFORM_DEFAULTS entry, so it
inherited the noisy global ('all') display defaults and narrated tool
progress / heartbeats / busy-ack detail into a permanent-message iMessage
thread. Register it as TIER_LOW alongside BlueBubbles and Signal, plus a
regression test guarding the tier.
Salvaged from #50511 (Photon tier piece only).
The PR's inline toolset resolution (checking 'memory' in cli_toolsets
list) produced wrong results for composite toolsets like 'hermes-cli'
which expand to include the memory tool. Replace with the canonical
_get_platform_tools() from tools_config.py which correctly handles
composite toolsets and all edge cases.
Update tests to mock _get_platform_tools instead of raw config.
The 'Built-in: always active' label was a hardcoded string that never
reflected the user's actual configuration. It now shows three separate
indicators, each reading from the real source of truth:
- Memory injection: reads memory.memory_enabled from config.yaml
- User profile: reads memory.user_profile_enabled from config.yaml
- Memory tool: checks if 'memory' is in platform_toolsets.cli
(or defaults to enabled if no explicit list)
Before:
Built-in: always active
After:
Built-in (MEMORY.md / USER.md):
Memory injection: disabled ✗
User profile: disabled ✗
Memory tool: disabled ✗
Invalid-tool exhaustion and truncated-tool early returns skipped finalize_turn, leaving role=tool transcripts that become tool→user on the next turn for strict providers. Call close_interrupted_tool_sequence before persist on those paths (same as interrupt aborts).
Hoist the duplicated check_info(source_id) call out of both
if/else branches into a single call after the branch. Remove
trailing whitespace on the blank line after the except block.
Follow-up cleanup for PR #69981.
Consolidate the two near-identical warning strings in
_log_wal_reset_bug_once into a single logger.warning call with an
action variable. Remove overengineered defensive tuple-length handling
in is_sqlite_wal_reset_vulnerable (sqlite3.sqlite_version_info always
returns a 3-tuple). Remove extra blank line.
Follow-up cleanup for PR #69981.
Assert the version matrix, fresh-DB DELETE fallback, already-WAL left
alone (no checkpoint/DELETE), fixed-SQLite WAL path, and warn-only
doctor output for vulnerable builds (#69784).
On vulnerable SQLite (e.g. 3.50.4), do not enable WAL for fresh/non-WAL
shared databases — prefer DELETE instead. Leave existing on-disk WAL
alone (no live downgrade under concurrent gateway/cron openers). Surface
Python/SQLite version details as a doctor warning (#69784).
Guard the globalThis gateway-state container and the survivor park/adopt
calls on the import.meta.hot literal (not the runtime hmrActive() helper),
so Vite dead-code-eliminates every HMR path in production. Prod now uses a
plain module-local singleton — no globalThis, no Symbol.for — and the
survivor module drops out of the bundle entirely. Verified: gatewayRegistryState,
gatewaySurvivor, and import.meta.hot are all absent from the prod build.
Removes the now-unused hmrActive() export.
Park the live primary gateway socket on Fast Refresh dispose and re-adopt it
on remount so dev UI edits don't tear down the WebSocket. Hold gateway store
singletons on globalThis + self-accept HMR on store/gateway.ts. Prod strips
import.meta.hot — live unmount unchanged.
Two fixes for desktop hands-free voice:
- The live speech session bound to the first assistant bubble with text, so
a tool-calling turn spoke only the opening narration and silently dropped
every later interim AND the final answer. The conversation selector now
aggregates all unspoken assistant bubbles in order (turn-scoped speech);
auto-speak keeps its latest-reply-only behavior.
- The speak-stream WS producer blocked forever on the text queue, so a
narration line with no trailing whitespace ("Let me check.") sat in the
sentence chunker until end-of-turn — spoken long after the tool finished,
with the UI stuck on "Preparing audio…". Mirror the CLI speaker's idle
flush: sentence-terminated buffers flush after 0.5s of producer silence,
anything else after ~2s; open <think> blocks are never flushed.
* fix(desktop): render agent credit notices as toasts (#69808)
The desktop renderer had no handler for the `notification.show` /
`notification.clear` WS events, so every credit-usage notice the backend
sends (`agent/credits_tracker.py` → `tui_gateway/server.py`) was silently
dropped. Credit warnings like "• Credits 50% used · $220.00 cap" never
appeared, even though the Ink TUI renders them in its status bar.
Add the two missing branches to the gateway-event dispatcher, delegating
to a small, pure-testable module:
- `store/agent-notices.ts` — `noticeToToast()` maps a notice to a toast
(level → toast kind, sticky → durationMs 0, ttl → ttl_ms), and uses the
notice `key` as the toast id. Re-emitting the same key REPLACES the
toast, so the credits 50→75→90 line escalates in place instead of
stacking, and a key-matched `notification.clear` maps straight to
`dismissNotification(key)`.
- The notice `text` already carries its own glyph (• ⚠ ✕ ✓), so no toast
icon is added.
- Notices are account-wide, so the toast shows regardless of which
session is focused.
The Ink TUI (`ui-tui/src/app/turnController.ts`) is the reference for the
latest-wins / sticky-vs-ttl / key-matched-clear behavior.
Export `NotificationInput` so the mapping's return type can be named.
* feat(desktop): native OS credit alerts + billing-page nudge (#69808)
Round out the credit-notice handling from the previous commit with the two
optional pieces from the issue:
- Native OS notification for the urgent pair. `credits.depleted` /
`credits.restored` also fire an Electron notification when Hermes is
backgrounded, via a new `credits` NativeNotificationKind (the existing
five didn't fit) with its own toggle in Settings → Notifications (the
panel is data-driven off NATIVE_NOTIFICATION_KINDS, so the toggle and
i18n are the only additions). The escalating usage line and grant-spent
notice stay in-app toasts only. Dispatch is `global` (account-wide, not
session-bound) and gated by the user's prefs + backgrounded check.
- Billing-page nudge. A `credits.*` crossing invalidates the
`['billing','state']` query so Settings → Billing reflects the change
immediately instead of waiting up to 30s for its poll.
`nativeNoticeInput()` is a pure mapping (urgent-key gate → native input),
unit-tested directly; the gateway-event branch does the localized-title
lookup and gated dispatch. i18n added for all four locales.
* feat(credits): report $used of $cap instead of % in the usage notice
The usage gauge is Nous subscription-cap-only (used_fraction requires a cap;
non-Nous providers emit no headers, so no notice fires). A bare percentage
implied a universal unit that doesn't exist, so report the absolute dollars
used of the cap instead: used = cap - remaining, from micros (money-safe),
clamped to [0, cap]. Still a snapshot at band-crossing (re-emits on band
change, not every turn) to keep the single escalating line and stay quiet on
append-only surfaces (messaging pushes one message per crossing).
* fix(desktop): de-dupe credit toast icon, band-color the figure, split detail
Three fixes to how agent credit notices render as toasts:
- Strip the leading severity glyph (the toast already draws a kind icon, so the
raw text doubled it). Native OS notifications keep the glyph (no icon there).
- Icon top-margin is now 0.42ch (font-relative) instead of a fixed rem.
- Band-color the $used figure (semibold) by $used/$cap: muted <75%,
--ui-orange >=75%, --ui-red >=90% (depleted red, restored green), reusing the
existing --ui-* usage palette. Icon shares the accent.
- Split a trailing '. detail' into a muted secondary line (title+description
convention) instead of an inline middot.
Generic 'accentColor' + 'meta' slots on the notification; degrades gracefully
when a notice has no figure.
* fix(desktop): billing page always fetches fresh state (team-account desync)
The x-nous-credits-* headers are best-effort and can drift out of sync,
notably in team/org accounts where another member's spend moves the shared
balance without touching this client's headers. The billing endpoint is the
source of truth, so the page no longer trusts a cache: staleTime 0 +
refetchOnMount 'always' force a fresh fetch on every open and focus (still
polling 30s while mounted). The credits.* invalidation nudge still pulls a
crossing in immediately.
* chore(desktop): dev-only credit-notice demo hotkey
Ctrl+Shift+C (and window.__creditsDemo()) steps the full credit-notice
lifecycle (usage 50->75->90, grant-spent, depleted/restored) through the real
gateway event fan-out via a new emitLocalGatewayEvent, so the toast/native/
billing-invalidation paths are testable without hitting real usage bands.
Installed only under import.meta.env.DEV, so it's tree-shaken from production.
Ctrl+Shift+C (and window.__creditsDemo()) steps the full credit-notice
lifecycle (usage 50->75->90, grant-spent, depleted/restored) through the real
gateway event fan-out via a new emitLocalGatewayEvent, so the toast/native/
billing-invalidation paths are testable without hitting real usage bands.
Installed only under import.meta.env.DEV, so it's tree-shaken from production.
The x-nous-credits-* headers are best-effort and can drift out of sync,
notably in team/org accounts where another member's spend moves the shared
balance without touching this client's headers. The billing endpoint is the
source of truth, so the page no longer trusts a cache: staleTime 0 +
refetchOnMount 'always' force a fresh fetch on every open and focus (still
polling 30s while mounted). The credits.* invalidation nudge still pulls a
crossing in immediately.
Three fixes to how agent credit notices render as toasts:
- Strip the leading severity glyph (the toast already draws a kind icon, so the
raw text doubled it). Native OS notifications keep the glyph (no icon there).
- Icon top-margin is now 0.42ch (font-relative) instead of a fixed rem.
- Band-color the $used figure (semibold) by $used/$cap: muted <75%,
--ui-orange >=75%, --ui-red >=90% (depleted red, restored green), reusing the
existing --ui-* usage palette. Icon shares the accent.
- Split a trailing '. detail' into a muted secondary line (title+description
convention) instead of an inline middot.
Generic 'accentColor' + 'meta' slots on the notification; degrades gracefully
when a notice has no figure.
The usage gauge is Nous subscription-cap-only (used_fraction requires a cap;
non-Nous providers emit no headers, so no notice fires). A bare percentage
implied a universal unit that doesn't exist, so report the absolute dollars
used of the cap instead: used = cap - remaining, from micros (money-safe),
clamped to [0, cap]. Still a snapshot at band-crossing (re-emits on band
change, not every turn) to keep the single escalating line and stay quiet on
append-only surfaces (messaging pushes one message per crossing).
Round out the credit-notice handling from the previous commit with the two
optional pieces from the issue:
- Native OS notification for the urgent pair. `credits.depleted` /
`credits.restored` also fire an Electron notification when Hermes is
backgrounded, via a new `credits` NativeNotificationKind (the existing
five didn't fit) with its own toggle in Settings → Notifications (the
panel is data-driven off NATIVE_NOTIFICATION_KINDS, so the toggle and
i18n are the only additions). The escalating usage line and grant-spent
notice stay in-app toasts only. Dispatch is `global` (account-wide, not
session-bound) and gated by the user's prefs + backgrounded check.
- Billing-page nudge. A `credits.*` crossing invalidates the
`['billing','state']` query so Settings → Billing reflects the change
immediately instead of waiting up to 30s for its poll.
`nativeNoticeInput()` is a pure mapping (urgent-key gate → native input),
unit-tested directly; the gateway-event branch does the localized-title
lookup and gated dispatch. i18n added for all four locales.
The desktop renderer had no handler for the `notification.show` /
`notification.clear` WS events, so every credit-usage notice the backend
sends (`agent/credits_tracker.py` → `tui_gateway/server.py`) was silently
dropped. Credit warnings like "• Credits 50% used · $220.00 cap" never
appeared, even though the Ink TUI renders them in its status bar.
Add the two missing branches to the gateway-event dispatcher, delegating
to a small, pure-testable module:
- `store/agent-notices.ts` — `noticeToToast()` maps a notice to a toast
(level → toast kind, sticky → durationMs 0, ttl → ttl_ms), and uses the
notice `key` as the toast id. Re-emitting the same key REPLACES the
toast, so the credits 50→75→90 line escalates in place instead of
stacking, and a key-matched `notification.clear` maps straight to
`dismissNotification(key)`.
- The notice `text` already carries its own glyph (• ⚠ ✕ ✓), so no toast
icon is added.
- Notices are account-wide, so the toast shows regardless of which
session is focused.
The Ink TUI (`ui-tui/src/app/turnController.ts`) is the reference for the
latest-wins / sticky-vs-ttl / key-matched-clear behavior.
Export `NotificationInput` so the mapping's return type can be named.
Guard the post-start set_agent_cursor_enabled on _session._started so
call_tool cannot re-enter session.start() (matches the start_session
lifecycle guard).
cua_driver_update_check now short-circuits to None when no driver
resolves; CI has none installed, so pin resolve_cua_driver_cmd in the
update-check and env-sanitization tests.
Cut steady-state Computer Use latency without changing default behavior
or waiting on cua-driver:
- Cap screenshots via set_config(max_image_dimension) on session start
(config: computer_use.max_image_dimension, default 1456)
- Cache aux-vision routing per (provider, model) so captures skip
repeated load_config()
- Add computer_use.capture_after_mode (default som) so users can opt
follow-ups down to ax (elements only) for speed
Auto-detect now disables the cursor overlay on darwin as well as
headless/WSL2 Linux. After start_session, also call
set_agent_cursor_enabled(false) when the policy is on so older drivers
without --no-overlay still tear the overlay down.
Co-authored-by: David Metcalfe <80915+DavidMetcalfe@users.noreply.github.com>
Print gh-image stdout and stderr on attachment failures, then replace the
pending inline-evidence marker in the PR review comment with an escaped
failure notice before preserving the failing workflow result.
The hermes-sweeper review #4701565902 (2026-07-15) flagged two
consistency issues in `_cua_driver_supports_no_overlay` and one
additive-config concern:
1. `cua-backend.py:260` — the `cua-driver --help` support probe
inherited the full parent environment. cua-driver is a third-party
binary; every other spawn site in this file (manifest probe at
`:214`, MCP spawn at `:697`, install probe at `:997`) uses
`_sanitize_subprocess_env(cua_driver_child_env())`. The `--help`
probe should match. This was a low-impact leak (only help output
exits), but inconsistency is the wrong default for a third-party
subprocess.
2. `cua_backend.py:238` — when the manifest returned a `command`
different from the input `driver_cmd` parameter (e.g. a relocated
executable at `/opt/relocated/cua-driver` while the system binary
is at `/usr/bin/cua-driver`), the support probe ran against
`_CUA_DRIVER_CMD` (the default) instead of the manifest-discovered
`command`. Two failure modes:
- The wrapper binary supports `--no-overlay` but the system binary
doesn't → probe returns False → overlay kept despite capability.
- The system binary supports `--no-overlay` but the wrapper doesn't
→ probe returns True → MCP spawn crashes on the unknown flag.
3. The original commit bumped `_config_version` 31→32 for an additive
default (`computer_use.no_overlay: None`). AGENTS.md specifies that
additive defaults in existing sections are handled by deep merge
and should NOT trigger a version bump. After cherry-picking onto
current `origin/main` (which is already at 33), the bump is
effectively dropped — resolved to main's 33.
Changes:
- Add `env=_sanitize_subprocess_env(cua_driver_child_env())` to the
`--help` subprocess (with the same import + rationale comment as
the manifest probe).
- Pass `driver_cmd=command` (or `driver_cmd=driver_cmd` for the
fallback path) into `_mcp_args_with_overlay_flag`, so the support
probe runs against the binary that will actually be launched.
Tests (3 new):
- `test_help_probe_passes_sanitized_env` — verifies `subprocess.run`
is called with an `env=` kwarg.
- `test_manifest_command_drives_support_probe` — verifies the probe
runs against the manifest command when it differs from the input
driver_cmd.
- `test_fallback_uses_input_driver_cmd_for_support_probe` — verifies
the fallback path (no command in manifest) uses the input
driver_cmd.
- `test_probe_distinguishes_support_between_binaries` — sanity check
that the lru_cache key on `driver_cmd` prevents cross-binary
cache leakage.
File-revert negative test confirmed all three of the new
"manifest/probe" tests are load-bearing: with the pre-fix code, they
fail (probe runs against the default binary instead of the resolved
one); with the fix, they pass. 20/20 tests in
`tests/computer_use/test_cua_no_overlay.py` green.
`TestMcpInvocationResolution` (8/8) still green.
Refs: sweeper review #4701565902
Address review feedback from cross-vendor review (Flash + GPT-OSS):
1. Auto-detect now checks for headless Linux (no DISPLAY), WSL2
(/proc/version contains 'microsoft'), instead of all Linux.
Desktop Linux with a compositor keeps the overlay.
2. Add _cua_driver_supports_no_overlay() that probes cua-driver --help
to check if the flag is supported. Older drivers (< 0.6.x) reject
unknown flags, so passing --no-overlay would crash the MCP spawn.
3. Update tests to cover headless vs desktop Linux, WSL2 detection,
version probe, and the unsupported-driver fallback path.
- Add autouse fixture to TestMcpInvocationResolution to disable
--no-overlay flag so existing tests assert baseline args
- Make test_config_load_failure_fails_safe and test_missing_section_enables
platform-aware (Linux auto-detect returns True, macOS/Windows False)
cua-driver's cursor overlay rendering loop can consume CPU indefinitely
when idle (#28152, #47032). On Linux/WSL2, the overlay serves no visual
purpose and the rendering path is the primary source of idle CPU usage.
Add computer_use.no_overlay config option (default: auto-detect) that
passes --no-overlay to cua-driver when enabled. Auto-detection disables
the overlay on Linux (covers WSL2, headless, containers) where it has no
benefit, and keeps it enabled on macOS/Windows where it is visually
useful.
Refs: #28152, #47032
Keep thin-PATH resolution while asserting the absolute binary on the
already-installed version check. Map contributor emails for the salvage.
Co-authored-by: Tianqing Yun <yuntianqing@yahoo.com>
Co-authored-by: Trevor Gordon <trevorbgordon@gmail.com>
Co-authored-by: Adrian Soto Mora <adrian.soto6@gmail.com>
Supersedes #52948 and #67402. Closes#52945.
Standalone hermes dashboard/serve was trusting HERMES_WEB_DIST and
HERMES_SERVE_HEADLESS inherited from a Desktop Electron parent, which
could serve the packaged desktop renderer ("Desktop IPC bridge is
unavailable") or disable the SPA. Drop only Electron-packaged WEB_DIST
paths (app.asar*) when HERMES_DESKTOP!=1, and clear inherited headless
for non-serve launches, while preserving caller-managed custom dist
overrides and the desktop-spawned backend path.
Co-authored-by: Bartok9 <danielrpike9@gmail.com>
Co-authored-by: Commander <commander@tianji.local>
* fix(desktop): improve fallback tool-call details
Show failed image-generation calls through the normal fallback row, remove duplicate normal-mode web-search JSON, and format Technical Mode payloads as readable JSON.
* feat(desktop): render terminal tool calls as transcripts
Show terminal commands with a prompt and exit status, then reveal ANSI-safe stdout and stderr in the expanded tool row.
* fix(desktop): reconcile tool calls by command
Match context-only tool starts with command-bearing completions when their IDs differ, preventing stale duplicate terminal rows. Show the web-search query above its result cards.
Adds the missing write path for the per-task model_override column (which
was previously only settable via manual SQL) and pairs it with a
provider_override so cross-provider switches resolve correctly:
- kanban_db: provider_override column (+migration), set_model_override()
with model_override_set event, create_task(model_override=,
provider_override=), dispatcher spawns worker with -m <model>
[--provider <name>]
- dashboard: Model row in the task drawer — dropdown fed by a new
/model-options endpoint (build_models_payload substrate, provider-grouped,
free-text fallback), PATCH + bulk model override support
- CLI: kanban create --model/--provider, new kanban set-model subcommand,
show prints the provider
- agent tools: kanban_create accepts model/provider; show/list expose
provider_override
Rate-limit recovery flow: override is settable on running tasks and takes
effect on the next dispatch, without touching the worker profile's config.
`resumeTile` — the cold-resume path for a session opened in a tile / split
pane — resumed with `{ session_id, cols }` and read messages with no profile,
so a tile opening a session from another profile let the gateway fall back to
the launch-profile DB and fork the conversation into the wrong profile: the
same cross-profile bleed the recovery resumes had (#67603), just a sibling
call path. Resolve the owning profile via the shared `resolveSessionProfile`
and carry it on both the transcript prefetch and the resume RPC.
Co-authored-by: oliviaaaa7788 <oliviaaaa7788@users.noreply.github.com>
Narrow the setup-error matcher, match the server's empty-key contract,
and route gateway-event plus create/resume/branch through one
credential-warning policy. Preserve configured=true only for
non-authoritative transport fallback.
Co-authored-by: Yingliang Zhang <zhangyingliang@outlook.com>
Co-authored-by: Brandon R <kingdomwarrior23@gmail.com>
`no-key-required` is a valid sentinel for local/self-hosted/custom
routers. Warning on it made Desktop treat a working setup as missing
API keys.
Co-authored-by: Yingliang Zhang <zhangyingliang@outlook.com>
Co-authored-by: Brandon R <kingdomwarrior23@gmail.com>
* fix(desktop): stop model-switch dup + route recovery resumes to the owning profile
Fixes two Desktop session-reconciliation symptoms from #67603.
Symptom 1 — duplicated user bubble after a model switch. The gateway
persists model-switch / personality notices as role=user `[System: …]`
rows (tui_gateway/server.py) so strict OpenAI-compatible providers don't
reject a non-leading system message (#48338). `preserveLocalPendingTurnMessages`
paired local optimistic rows with the stored transcript by user-role
ordinal, so a marker between two real user turns shifted every later
ordinal and the optimistic row was re-appended at the bottom. The single
trailing-marker case is already covered by the compression-era
`latestAuthoritativeUser` guard, but two switches around one turn (marker
before AND after the committed prompt) still duplicated it. Exclude
`[System:` bookkeeping markers from ordinal pairing on both sides.
Symptom 2 — a session appearing under two profiles. The main resume path
already resolves a session's owning profile via `resolveStoredSession`
(cache → active backend → cross-profile probe), but the recovery
`session.resume` calls (stale runtime id, session-not-found, wedged loop,
redirect) omitted `profile`, so the gateway fell back to the launch-profile
DB and forked the conversation into the wrong profile. Route every recovery
resume — and an uncached right-click branch — through the same resolver so
the profile is carried even for sessions outside the paginated sidebar
window (the cache-miss gap).
Tests: discriminating two-switch marker test (fails before, passes after);
cache-hit + cross-profile cache-miss coverage for the recovery resume and
for branching an uncached session.
Supersedes #68665 and #63590.
Closes#67603.
Co-authored-by: Dolverin <5910064+Dolverin@users.noreply.github.com>
Co-authored-by: oliviaaaa7788 <274182427+oliviaaaa7788@users.noreply.github.com>
* fix(desktop): scope the remembered session id per profile
A single global `hermes.desktop.lastSessionId` key remembered ONE session
across every profile, so a relaunch or cold start under profile B would try
to restore a session owned by profile A — reinforcing the impression that a
conversation had bled between profiles (#67603, second symptom).
Key the remembered id by the session's owning profile (resolved from the
session row's `profile`, falling back to the active gateway profile), read it
back for the active profile on restore, and clear an exhausted session under
its owner. The default profile keeps the original unsuffixed key so existing
installs' remembered session survives the upgrade.
Co-authored-by: oliviaaaa7788 <oliviaaaa7788@users.noreply.github.com>
* fix(gateway): hide [System:] bookkeeping markers from every transcript
Model-switch and personality notices are persisted as role=user `[System: …]`
rows so strict providers accept them mid-history, but they are model-facing
runtime metadata, not user turns. `_history_to_messages` — the single display
projection every client reads — passed them straight through, so on resume or
reload they rendered as a fake user bubble in the desktop, TUI, CLI, and web
transcripts.
Drop them in that projection. The raw marker stays in `session["history"]`
for the model, so nothing changes for inference; only the display loses a row
that never belonged to the user. This also removes the stored marker from the
payload the desktop reconciles against, killing the ordinal shift that
duplicated the optimistic prompt (#67603) at its source — the desktop-side
marker exclusion remains as a fallback for older backends.
Co-authored-by: Dolverin <Dolverin@users.noreply.github.com>
---------
Co-authored-by: Dolverin <5910064+Dolverin@users.noreply.github.com>
Co-authored-by: oliviaaaa7788 <274182427+oliviaaaa7788@users.noreply.github.com>
Co-authored-by: oliviaaaa7788 <oliviaaaa7788@users.noreply.github.com>
Co-authored-by: Dolverin <Dolverin@users.noreply.github.com>
* fix: serialize TUI gateway websocket sends
* fix(tui_gateway): preserve websocket batch order
* refactor(tui_gateway): drop unused _safe_send wrapper
The batch-serialization fix routes every send through _safe_send_many;
_safe_send became a dead single-line wrapper with no callers. Remove it.
---------
Co-authored-by: supplefrog <78985073+supplefrog@users.noreply.github.com>
Co-authored-by: Brooklyn Nicholson <brooklyn.bb.nicholson@gmail.com>
Post-rebase consolidation over the merged C7/C16/C10 work:
- _ensure_dm_conversation now records workspace ownership via
_remember_channel_team and bounds _dm_conversation_cache (cap 5000)
- module-level _slack_dm_cache (C7 standalone path) bounded oldest-first
- _user_is_bot_cache (C10) bounded with _trim_oldest_dict_entries
- caption-mode contract tests updated for the C7+C8 merged media path
Widen the #19237 send_message fix to the live adapter: send, _upload_file,
send_multiple_images, send_image, send_video, send_document,
send_exec_approval, send_slash_confirm, and send_clarify now route bare
Slack user IDs (U.../W...) through a shared _ensure_dm_conversation helper
before calling chat.postMessage / files_upload_v2, which reject user IDs.
This closes the gap in #17261 where an attachment worked when replying in
a thread but failed when directed at a user DM, and extends the DM-open
fallback to clarify/approval Block Kit prompts so gated actions can reach
a user directly.
Resolution uses the workspace-scoped client (multi-workspace installs open
the DM with the right bot token), caches per (team, user), and records the
opened D... channel in the channel→team map. On failure the original
target passes through so the downstream API call surfaces the real Slack
error.
Fixes#17261
Refs #19236
The post-stream helper (_deliver_media_from_response) rescanned the
already-streamed response and promoted bare local filesystem paths into
real uploads via extract_local_files. Since the visible reply was already
streamed verbatim, any bare path there is either text the user has seen
or stale inspected/tool content — not an attachment request. On Slack this
uploaded images from stale inspected content after otherwise clean replies.
Post-stream delivery now honors only explicit MEDIA: directives. The
non-streaming path in gateway/platforms/base.py keeps its bare-path
auto-detect, because that path controls the visible text and strips the
path from the reply when it attaches — auto-attach is intentional there.
Regression tests: bare image/document paths in a streamed reply produce
no upload; explicit MEDIA: tags still deliver.
Fixes#20834
Add early auth check in _handle_slack_message() that runs BEFORE any
API calls (thread context fetch, user name resolution, file downloads)
or file processing. Unauthorized users could previously trigger Slack
API calls and file downloads before the runner's _is_user_authorized
gate rejected them.
Same pattern as Telegram fix#54164: build a SessionSource and check
the runner's _is_user_authorized at the adapter level before event
construction consumes resources.
Fixes the gap where Slack has no adapter-level auth gate while Discord
and Telegram have adapter-level allowlists.
Slack could already deliver files in-channel via the gateway, but
send_message omitted MEDIA for Slack and told the model it was
unsupported — causing agents to inconsistently refuse PDF sends.
Wire Slack through files_upload_v2 in the standalone sender.
Model-switch and personality notices are persisted as role=user `[System: …]`
rows so strict providers accept them mid-history, but they are model-facing
runtime metadata, not user turns. `_history_to_messages` — the single display
projection every client reads — passed them straight through, so on resume or
reload they rendered as a fake user bubble in the desktop, TUI, CLI, and web
transcripts.
Drop them in that projection. The raw marker stays in `session["history"]`
for the model, so nothing changes for inference; only the display loses a row
that never belonged to the user. This also removes the stored marker from the
payload the desktop reconciles against, killing the ordinal shift that
duplicated the optimistic prompt (#67603) at its source — the desktop-side
marker exclusion remains as a fallback for older backends.
Co-authored-by: Dolverin <Dolverin@users.noreply.github.com>
A single global `hermes.desktop.lastSessionId` key remembered ONE session
across every profile, so a relaunch or cold start under profile B would try
to restore a session owned by profile A — reinforcing the impression that a
conversation had bled between profiles (#67603, second symptom).
Key the remembered id by the session's owning profile (resolved from the
session row's `profile`, falling back to the active gateway profile), read it
back for the active profile on restore, and clear an exhausted session under
its owner. The default profile keeps the original unsuffixed key so existing
installs' remembered session survives the upgrade.
Co-authored-by: oliviaaaa7788 <oliviaaaa7788@users.noreply.github.com>
Fixes two Desktop session-reconciliation symptoms from #67603.
Symptom 1 — duplicated user bubble after a model switch. The gateway
persists model-switch / personality notices as role=user `[System: …]`
rows (tui_gateway/server.py) so strict OpenAI-compatible providers don't
reject a non-leading system message (#48338). `preserveLocalPendingTurnMessages`
paired local optimistic rows with the stored transcript by user-role
ordinal, so a marker between two real user turns shifted every later
ordinal and the optimistic row was re-appended at the bottom. The single
trailing-marker case is already covered by the compression-era
`latestAuthoritativeUser` guard, but two switches around one turn (marker
before AND after the committed prompt) still duplicated it. Exclude
`[System:` bookkeeping markers from ordinal pairing on both sides.
Symptom 2 — a session appearing under two profiles. The main resume path
already resolves a session's owning profile via `resolveStoredSession`
(cache → active backend → cross-profile probe), but the recovery
`session.resume` calls (stale runtime id, session-not-found, wedged loop,
redirect) omitted `profile`, so the gateway fell back to the launch-profile
DB and forked the conversation into the wrong profile. Route every recovery
resume — and an uncached right-click branch — through the same resolver so
the profile is carried even for sessions outside the paginated sidebar
window (the cache-miss gap).
Tests: discriminating two-switch marker test (fails before, passes after);
cache-hit + cross-profile cache-miss coverage for the recovery resume and
for branching an uncached session.
Supersedes #68665 and #63590.
Closes#67603.
Co-authored-by: Dolverin <5910064+Dolverin@users.noreply.github.com>
Co-authored-by: oliviaaaa7788 <274182427+oliviaaaa7788@users.noreply.github.com>
- Replace undefined CSS class with Tailwind v4's
so long choice text wraps properly
instead of overflowing the button container.
- Add validation that strips non-string items,
blanks, newlines, and text >200 chars — preventing garbage/JSON
arrays from rendering as raw values.
- Add diagnostic logging at both the gateway event
handler and the tool-args parser when choices are dropped, so
malformed payloads are no longer silent.
- Apply in both the gateway event handler
() and the inline tool-args parser
() for consistent defense in depth.
- Add unit tests for covering null, non-array,
mixed-type, blank, multiline, and overlong inputs.
Closes#69122
Co-authored-by: webtecnica <webtecnica@users.noreply.github.com>
* fix(desktop): keep clarify lifecycle when tool progress is off
* fix(desktop): render clarify prompt from the request event
Re-authored onto the current use-message-stream/gateway-event.ts (the
original patched the pre-split use-message-stream.ts). When the tool.start
row that normally mounts the inline clarify UI is missed (stream reconnect
/ hydration race), upsert a stable pending clarify tool row from
clarify.request itself so the prompt stays answerable; a real
tool.start/complete with the same request id merges rather than duplicates.
Co-authored-by: 정수환 <centerid@naver.com>
* chore(contributors): map centerid@naver.com -> lidises
Attribution mapping for the salvaged #47544 commit.
* fix(desktop): correlate clarify rows by question so hydration can't duplicate
The hydrated row (from clarify.request's request_id) and the real tool.start
row (the model's tool_call_id) have different ids, so id-only matching appended
a second clarify card in the normal path (caught by the BLOCKING_CLARIFY e2e:
'question' resolved to 2 elements). Add 'question' to the tool match-value keys
so a clarify upsert merges into the existing pending clarify row regardless of
id (same request<->args correlation ClarifyToolPending already uses); when no
row exists yet (reconnect/hydration) it still creates one.
---------
Co-authored-by: 정수환 <centerid@naver.com>
* fix(gateway): deliver assistant prose before clarify poll
The clarify poll is sent on a separate, agent-thread-blocking path while
buffered assistant prose (interim commentary / streamed deltas) sits in
the GatewayStreamConsumer queue, drained asynchronously. The poll won the
race, so the question rendered ABOVE its own explanation, and a redundant
'clarify: ...' tool-progress bubble wedged between them.
- Add GatewayStreamConsumer.flush_pending_sync(): a synchronous flush
barrier (_FLUSH sentinel + threading.Event) that blocks the agent
thread until everything queued before it is finalized and delivered.
- Call it in the gateway clarify callback before send_clarify, so prose
always lands before the poll. Best-effort with a 3s timeout.
- Suppress the redundant clarify tool-progress bubble (the poll already
shows the question + options).
Tests: 3 new ordering/timeout cases in test_stream_consumer.py.
(cherry picked from commit 9a6e27badb7646c839902db1d2f4a8affcf38b1e)
* chore(contributors): map matvey.sakhnenko03@icloud.com -> sakhnenkoff
Attribution mapping for the salvaged #54328 commit (Cluster C).
---------
Co-authored-by: Matvii Sakhnenko <matvey.sakhnenko03@icloud.com>
* test(clarify-gateway): cover signature, timeout fallback, and notify paths for 100% coverage
Fixes#36531
(cherry picked from commit 5265dfe2f5d484c5fb97c71eb311cbb230c0110e)
* fix(clarify): one canonical timeout across CLI, TUI/desktop, and gateway
The clarify wait timeout was resolved three different (wrong) ways:
- CLI (`cli.py`, `hermes_cli/callbacks.py`) read a non-existent top-level
`clarify.timeout`, so it always fell through to a hardcoded 120s instead of
the canonical `agent.clarify_timeout` (default 3600) the gateway uses (#42969).
- The TUI/desktop bridge called `_block("clarify.request", …)` with no timeout,
so it used the hardcoded 300s `_block` default and ignored config (#51960).
- There was no way to disable the auto-skip: a user who wanted the agent to wait
indefinitely while they think couldn't get it.
Collapse all of this onto a single resolver:
- `tools.clarify_gateway.resolve_clarify_timeout(config)` is the one source of
truth. Order: explicit legacy `clarify.timeout` (back-compat) → canonical
`agent.clarify_timeout` → 3600. `<= 0` is preserved verbatim as "unlimited".
- CLI, callbacks, and the TUI bridge (`_clarify_timeout_seconds`) all route
through it, so the three surfaces can't drift.
- `<= 0` means unlimited everywhere: `wait_for_response` and `_block` drop the
deadline (heartbeat still fires), and the CLI hides its countdown.
Tests: resolver order / default / non-numeric / unlimited-sentinel; an
unlimited `wait_for_response` blocks until resolved rather than auto-skipping;
the TUI clarify bridge passes the configured timeout to `_block`.
Supersedes #42974 (CLI key), #51993 (TUI honors config), and #68986 (unlimited
wait); folds in #52031 (clarify_gateway coverage).
Co-authored-by: liuhao1024 <liuhao1024@users.noreply.github.com>
Co-authored-by: lkevincc0 <lkevincc0@users.noreply.github.com>
Co-authored-by: theone139344 <theone139344@users.noreply.github.com>
Co-authored-by: baauzi <baauzi@users.noreply.github.com>
---------
Co-authored-by: Christopher-Schulze <210261288+Christopher-Schulze@users.noreply.github.com>
Co-authored-by: liuhao1024 <liuhao1024@users.noreply.github.com>
Co-authored-by: lkevincc0 <lkevincc0@users.noreply.github.com>
Co-authored-by: theone139344 <theone139344@users.noreply.github.com>
Co-authored-by: baauzi <baauzi@users.noreply.github.com>
`_block()` bridges four blocking request types — secret, sudo, clarify,
terminal.read — with an identical lifecycle: on timeout the tool gives up and
returns empty, but a slow renderer (or a WebSocket reconnect that dropped
tool.complete) can still answer afterward. Only secret and sudo tolerated that
late reply; clarify and terminal.read still hit the generic 4009 "no pending
request" error, which clients surface as a raw JSON-RPC string (and at least
one desktop fork re-armed the pending request on the error).
Bring the two stragglers in line with the pair that already works:
- `_block` now emits `{event}.expire` on timeout for all four request types.
- `clarify.respond` and `terminal.read.respond` pass `allow_expired=True`, so a
late answer resolves to `{"status": "expired"}` instead of erroring.
Tests parametrize the timeout-expiry and late-idempotent-response cases over
all four bridges; the old test asserting clarify stays a 4009 error is updated
to the new graceful contract.
Supersedes #56571 (clarify) and #64886 (terminal.read) — same root cause, one fix.
Co-authored-by: liuhao1024 <liuhao1024@users.noreply.github.com>
Co-authored-by: pierrenode <pierrenode@users.noreply.github.com>
`ExpandableBlock` used a single full-width `absolute inset-x-0 bottom-0`
button that was BOTH the fade overflow cue and the expand/collapse control.
It sat on top of the whole 28px bottom strip, so it swallowed:
- the horizontal scrollbar of a wide code block (couldn't scroll sideways), and
- pointer/drag events on the block's last line (couldn't select/copy the tail).
Split the two responsibilities: the full-width fade stays as a pure
`pointer-events-none` cue, and the only clickable target is a compact toggle
pinned to the right edge (`pointer-events-auto`), clear of the draggable
scrollbar track. The inner container gains `overflow-x-auto` so wide code
gets a working scrollbar. Applies uniformly to code cards and the plain-text
fallback — no per-call prop needed.
Regression test asserts the pointer-events contract (fade `pointer-events-none`
+ full-width, toggle `pointer-events-auto` + right-pinned, not `inset-x-0`) and
that the toggle still flips `aria-expanded`.
Supersedes #69558, #69428Fixes#69168
Co-authored-by: alelpoan <alelpoan@proton.me>
Co-authored-by: BerneYue <xiongyue_hnu@163.com>
Classic /model confirmation already threaded custom_providers into the
context display, but the shrink-warning path did not. Probe-down then
matched the hardcoded "qwen" catalog (131072) and falsely warned that a
1M custom endpoint had shrunk — while the status bar still showed 1M.
Pass the same fresh inventory list used by switch_model/TUI (with
agent-snapshot fallback), and fall back to agent._custom_providers inside
merge_preflight_compression_warning when the kwarg is omitted.
Limit the _NET_ACTIVE_WINDOW xprop fallback to unqualified default
captures so exact pid/window_id targeting does not pay up to a 2s
subprocess probe on Linux/X11.
When Linux/X11 reports the same z_index for every on-screen window,
prefer _NET_ACTIVE_WINDOW via xprop instead of list order. Keep the
higher-z-index-is-frontmost contract when ordering is informative.
- Add SLACK_THREAD_REQUIRE_MENTION, SLACK_IGNORE_OTHER_USER_MENTIONS and
SLACK_REQUIRE_MENTION_CHANNELS to the conftest behavioral-env force-clear
list so config-loader side effects can't leak between tests (same class
of leak the existing SLACK_* entries guard against).
- Document thread_require_mention and require_mention_channels in the
Slack messaging guide next to the other mention-gating options.
Port of the Slack half of #13855 (by @kshitijk4poor), reimplemented against
the plugin adapter (the original PR targets the deleted
gateway/platforms/slack.py and six other legacy adapters).
Channels listed in slack.require_mention_channels (config.yaml) or
SLACK_REQUIRE_MENTION_CHANNELS ALWAYS require an explicit @mention, even
when require_mention is false globally or the channel is in
free_response_channels — the opposite direction of free_response_channels.
Instead of duplicating the PR's inline reply-to-bot-thread/mentioned-thread/
session checks, the forced channel falls through to the SAME decision chain
as normal mention gating, so all five wake checks in
_should_wake_on_unmentioned_message keep applying (single decision path).
Credit: adapted from #13855 by @kshitijk4poor (Slack half only; the other
platform halves target deleted legacy adapters and are out of scope for
this cluster).
Adds the option to the Mention & Trigger Behavior section: leading-
mention semantics, opt-in default, env var, and DM/MPIM scope (1:1 DMs
unaffected; MPIMs apply it like channels).
Claude-Session: https://claude.ai/code/session_01TKsNdptNdo9CqT2u7JMdkH
The ignore_other_user_mentions gate relied solely on is_mentioned, which
only recognises exact <@UID> markup, so a message mentioning the bot in
pipe form (<@UID|name>) alongside a leading other-user mention was
wrongly suppressed. The gate now also scans for the bot's own mention in
either markup before ignoring a message.
Claude-Session: https://claude.ai/code/session_01TKsNdptNdo9CqT2u7JMdkH
Once the bot is @mentioned in a Slack thread it auto-follows and replies
to every later message in that thread, including ones a human addresses to
another human (e.g. "@rasha check this out"). The bot butts in.
Add slack.ignore_other_user_mentions (env SLACK_IGNORE_OTHER_USER_MENTIONS,
default off). When on, a channel/thread message whose first token @mentions
someone other than the bot is treated as addressed to that person and the
bot stays silent unless it is also mentioned. This is Slack parity for the
Discord option of the same name (#33501), adapted to Slack's thread model:
the trigger is a leading mention ("addressed to"), so a message that merely
references another user mid-sentence still reaches the bot.
The gate sits ahead of the free-response / require_mention ladder so it also
overrides the mentioned-thread auto-follow. DMs are never filtered.
Follow-up rework on the #51627 cherry-pick:
- Guard the 5th wake check (parent-mentioned-bot, #24848) against a None
parent_text — _fetch_thread_parent_text is typed to return str but tests
(and defensive callers) can surface None; 'in None' raised TypeError.
- Drop the PR's fixture-level _fetch_thread_parent_text/_fetch_thread_context
AsyncMocks from TestThreadReplyHandling/TestAssistantThreadLifecycle: main's
#24848 tests exercise the real parent-text path via conversations_replies
side effects, and the blanket mocks broke them.
- _resolve_user_is_bot reworked to the workspace-scoped (team_id, user_id)
cache key introduced by the multi-workspace name cache on main.
Keep Enter as the Cursor-style live-turn steering gesture and assign
Ctrl/Cmd+Enter to queue the current draft. Align keybind settings,
tooltips, translations, and the Electron queue-boundary coverage.
Widening pass over the whole adapter following the cluster-C16 audit
(#51019, #51097, #23676, #23375): every in-memory structure that
accumulates per-message, per-user, or per-thread state is now bounded,
and every eviction is oldest-first — never arbitrary set-iteration
order, which is the #51019 failure mode (bot silently going quiet on
the most ACTIVE thread because set.pop-order eviction removed it).
Newly bounded:
- _approval_resolved / _clarify_resolved (caps 1000): unclicked
approval/clarify prompts leaked their double-click-guard entries
forever; oldest-insertion eviction via _trim_oldest_dict_entries.
- _reacting_message_ids (cap 5000): reaction lifecycle entries leaked
when an exception fired between add and finalize; oldest-ts eviction.
- _active_status_threads (cap 1000): statuses abandoned by error paths
accumulated; oldest-thread-ts eviction so the newest live status is
never cleared.
- _channel_team (cap 10000): grew with every DM channel the bot ever
saw (DM channel IDs are per-user). All four write sites now route
through _remember_channel_team; eviction is safe because entries are
re-learned from the next event and _get_client falls back to the
primary client.
- _slash_command_contexts (cap 1000): TTL cleanup only ran on lookup,
so contexts whose ephemeral replies never happened accumulated;
overflow purges expired entries first, then oldest-stash-first.
Converted from arbitrary set-order eviction to oldest-first:
- _titled_assistant_threads: keys are (team, channel, thread_ts) —
now evicts oldest thread first via _discard_oldest_by_thread_ts.
- _thread_rehydration_checked: keys are team:channel:thread_ts[:user] —
arbitrary eviction here would re-run an ACTIVE thread's restart
rehydration check and re-inject the missed-delta context; now evicts
oldest thread first.
- _reacting_message_ids uses #51097's _discard_oldest_slack_timestamps.
Deliberately NOT bounded (naturally tiny, per-workspace):
_team_clients, _team_bot_user_ids, _team_bot_names (one entry per
installed workspace), _assistant_threads / _agent_view_contexts /
_bot_message_ts / _mentioned_threads / _user_name_cache /
_thread_context_cache (already bounded), _dedup (MessageDeduplicator
has max_size + TTL internally).
New helpers: _trim_oldest_dict_entries (dicts preserve insertion
order, so oldest-first is exact) and _discard_oldest_by_thread_ts
(chronological sort on the embedded Slack ts for keyed sets).
Tests: caps hold under churn, eviction removes OLDEST not arbitrary
entries, newest/active entries survive eviction pressure (regression
shape for #51019), plus end-to-end paths through _resolve_user_name
and _handle_slash_command.
Part of the C16 cache-bounds consolidation with #51097 (markoub),
#23676 and #23375 (EloquentBrush). Fixes#51019.
_fetch_thread_context() caches per-thread Slack history under a
60-second TTL, but expired entries are never removed. On the current
adapter this is worse than when first reported: each entry now also
retains the raw conversations.replies payloads (messages) for
watermark re-formatting, so a busy multi-channel workspace accumulates
full message lists forever. _bot_message_ts, _mentioned_threads, and
_assistant_threads all enforce a MAX constant with active eviction in
the same __init__; _thread_context_cache had the TTL declared but no
matching eviction path.
Fix: add _THREAD_CACHE_MAX = 2500 and purge expired entries
(fetched_at older than _THREAD_CACHE_TTL) whenever a new write pushes
the cache past the limit. TTL-based eviction is used instead of LRU
because fresh entries are still needed; evicting them would
immediately re-trigger a rate-limited conversations.replies call.
Adds two unit tests: one verifying stale entries are purged on
overflow, one verifying fresh entries survive.
Reapplied from #23375 onto the current adapter (moved to
plugins/platforms/slack/adapter.py; cache entries now carry raw
message payloads and parent_user_id).
_resolve_user_name() caches every resolved (team_id, user_id) → display
name but never evicts entries. On a long-running bot in a large
workspace every unique user the bot encounters adds a permanent entry.
Sibling structures _bot_message_ts (BOT_TS_MAX=5000) and
_assistant_threads (ASSISTANT_THREADS_MAX=5000) already have caps;
_user_name_cache had none.
Fix: add _USER_NAME_CACHE_MAX = 5000 and evict the oldest half on
overflow after each write, matching the existing sibling pattern.
Consolidates the two cache-write branches (success + API error) into a
single write so eviction runs once per resolution.
Reapplied from #23676 onto the current adapter (cache moved to
plugins/platforms/slack/adapter.py and is now keyed by
(team_id, user_id) tuples for multi-workspace safety).
Companion to gateway-gateway#160 (connector absent-row default flips to
mention-gated / requireAddress=true).
relay_relevance_policy() previously returned None whenever the projected
policy was all-falsy, on the premise that 'the connector's quiet default
already matches'. Once the connector defaults requireAddress to true,
that premise inverts for one case: an agent EXPLICITLY configured with
require_mention: false would stay silent and get mention-gated anyway,
with no way out.
The nothing-to-declare check now keys on 'require_mention is unset'
rather than 'require_mention is falsy': an explicit false IS a
configured knob and is declared to the connector; a genuinely
unconfigured platform still declares nothing and inherits the
connector's default. No wire/contract change.
Cron jobs targeting a Slack user (deliver=slack:U…) failed with
channel_not_found: chat.postMessage and files_upload_v2 require a
conversation ID, and a DM must be opened first via conversations.open
to obtain a D… ID. The tool-level resolution in send_message() does not
cover cron's direct _send_to_platform → standalone_sender_fn path.
Add _resolve_slack_user_dm to the Slack plugin: resolves U…/W… targets
via conversations.open (proxy-aware, cached per token+user so repeated
cron fires don't re-open the DM), wired into _standalone_send ahead of
both the text and media delivery branches so every standalone entry
point benefits.
Adapted from #17444 by @Esther-Zhu023 — the original patched the legacy
_send_slack helper in tools/send_message_tool.py, which has since moved
into the plugin as _standalone_send (#41112).
Closes#17444
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Cron/async deliveries (reply_to=None) should always go to the home/target
channel, but _resolve_thread_ts() returns metadata.thread_id/thread_ts even
when reply_to is None, routing messages to the thread where the cron job
was created instead of the configured home channel.
Add early return None when reply_to is None, so async deliveries never
inherit stale thread context from metadata.
Closes#59097
The in_channel flat-delivery clear was gated on `runtime_adapter is not
None`, but the flat continuation session is seeded ONLY on the live-send
path, which also requires a running event loop. When an adapter is
present but the loop is absent/not-running, the live-send/seed block is
skipped and delivery falls through to the standalone path — clearing
thread_id there flattened an unseeded brief with no continuable session
behind it (and bypassed the D6 capability check).
Factor the live-send condition into `live_adapter_ready` and gate both
the thread_id clear and the live-send block on it, so the clear can no
longer drift from the seed. Add an adapter-present/no-running-loop
regression test (verified it fails against the un-scoped clear).
Addresses review r3609147550.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
cron_continuable_surface=in_channel is meant to deliver a cron brief FLAT
into a channel (thread_id=None) so a plain channel reply continues the job
via the shared-channel session. The scheduler already seeds that flat
session (_seed_cron_channel_session) but never cleared the ORIGINAL
target/origin thread_id before routing the actual delivery, so a job
scheduled from inside a Slack thread still delivered into that origin
thread — the seeded continuable session then never matched where the brief
actually landed.
Clear thread_id once in_channel_surface is finalized, before it is re-read
for route_thread_id / the standalone fallback. Scope the clear to EXACTLY
the target that gets flat-seeded — the origin-continuable target
(mirror_this_target) on a live, in_channel-capable adapter
(runtime_adapter is not None):
- fan-out / broadcast / explicit-thread targets keep their thread_id — they
are not continuable and are never seeded, so flattening them would
reroute an explicit thread and could collapse two targets into duplicate
flat sends;
- the standalone no-live-adapter path keeps the origin thread — it can
never seed the flat session, and without an adapter the D6 capability
fail-safe can't run, so flattening there would both drop the brief out of
any continuable lane and bypass D6.
This keeps the clear in lockstep with the seed condition
(in_channel_surface and mirror_this_target) and the D6 fail-safe.
mirror_this_target / origin_user_id are computed earlier using the original
thread_id to match the origin conversation.
Regression tests in TestCronContinuableSurfaceInChannel:
- a job whose origin carries a thread_id must not forward that thread_id to
the live adapter (DeliveryRouter folds target.thread_id into send
metadata), and the seeded continuable session stays flat (thread_id=None);
- with no live adapter, the standalone send falls back to the origin thread
rather than flattening.
Per review (Victor): response_url failure does not mean ephemeral delivery
is impossible — chat.postEphemeral is an independent API path that keeps
the reply private. Public-channel fallback removed entirely; when both
ephemeral paths fail the reply is dropped with a logged error rather than
leaked to the channel. No config knob needed.
- test_thread_command_skips_context_prefix: post-#69320 thread context IS
fetched on first thread entry but rides channel_context; assert the
command token stays at char zero and the backfill is preserved, instead
of asserting the fetch never happens.
- test_slack_send_retry.py: main's _get_client() now takes team_id;
update the lambda stubs.
Two silent-loss modes in the ephemeral slash reply path:
1. Delivery failure was swallowed: _send_slash_ephemeral returned
success=True on any POST failure, so the user's actual command reply
vanished behind the stale 'Running /cmd…' ack. It now returns
success=False and send() falls back to normal channel delivery.
2. Long replies were truncated to the first ~39k chunk with no notice.
Replies are now chunked across response_url POSTs (first replaces
the ack, follow-ups append, all ephemeral), capped at Slack's 5-POST
response_url budget with an explicit truncation notice when exceeded.
Also updates the #55357 bounded-error-read test for the #26788 precise
context matching (ContextVar must be set) and channel fallback.
Fixes#19688
Commands typed in Slack could be mangled by every enrichment layer the
adapter applies to normal messages:
- Block Kit / unfurl / attachment-notice / text-file injection could
prepend or append content around a command, moving the command token
away from character zero or polluting its arguments. Commands are now
restored from canonical authored input after all enrichment
(final is_command_text guard before MessageEvent construction).
- @bot /cmd (typed slash behind a mention) was never classified as a
command; the mention-strip branch now re-probes for both slash and
bang forms.
- The Slack Agent-view context label ([Slack app context: ...]) was
prepended to command events too; now command-exempt.
- Native slash payload arguments were strip()ed, destroying meaningful
spacing inside/after arguments; only the command delimiter is
nonsemantic now.
- Slash payload thread identity (thread_ts/message_ts, top-level or
nested in message/container) is preserved onto SessionSource so
session-scoped commands hit the same thread session.
- /queue and /steer queued fallbacks now propagate channel_context so a
command that triggered first-entry thread backfill doesn't lose the
history when re-queued.
Adapted from PR #66310 to the post-#69320 channel_context design (thread
history already rides MessageEvent.channel_context, never text).
Slack's send() caught all exceptions and returned a bare
SendResult(success=False) — never setting retryable=True or extracting
the server's Retry-After header. When Slack returned a 429 rate-limit
error, the base _send_with_retry() layer saw retryable=False and did
not retry, silently dropping remaining message chunks.
Reuse the existing _is_retryable_upload_error() helper (which already
detects 429, 500+, and connection-type errors) to set retryable=True,
and extract the Retry-After header from the SlackApiError response
when present so the base retry layer honors Slack's backoff schedule
instead of its own default.
Sibling of the Telegram FloodWait fix (PR #46762 / commit 404b06ac4)
which added the SendResult.retry_after plumbing to the base layer.
Adds five regression tests covering 429 with/without Retry-After,
500 server errors, 403 non-retryable errors, and connection errors.
_pop_slash_context fell back to a channel-only scan when the
_slash_user_id ContextVar was unset (i.e. send() invoked from a
non-slash code path such as a cron delivery or a normal channel reply).
That scan could steal another user's pending slash reply context: the
normal message got swallowed into an ephemeral response_url POST that
replaces the invoker's ack, and the slash invoker's actual reply then
posted publicly. Remove the fallback — when the ContextVar is unset,
match nothing.
Surgical reapply of PR #26788 (originally against gateway/platforms/slack.py).
_has_active_session_for_thread() hardcoded chat_type='group', causing
session key mismatch for DM and MPIM threads. DM sessions key as
agent:main:slack:dm:{chat_id}:{thread_ts} but the lookup built
agent:main:slack:group:{chat_id}:{user_id}:{thread_ts}.
Impact: _has_active_session_for_thread always returned False for DM
threads, causing thread context to be prepended on every message. The
prepended context broke slash command detection (get_command() checks
text.startswith('/')), so /cmd and !cmd never worked in DM threads.
Fix: accept event-derived chat_type parameter instead of hardcoding
'group'. Both call sites pass chat_type='dm' if is_dm else 'group',
where is_dm is already computed from channel_type in {'im', 'mpim'}.
This correctly handles:
- IM channels (D-prefix): chat_type='dm'
- MPIM channels (G-prefix): chat_type='dm' (was missed by D-prefix heuristic)
- Channel messages (C-prefix): chat_type='group' (unchanged)
Added regression tests covering DM thread lookup, MPIM thread lookup,
and negative cases verifying the old hardcoded 'group' behavior fails.
Slack rich_text blocks mirror the original message text. When bang
commands are rewritten from !model to /model, appending block text makes
the command arguments include a duplicate payload, so the model switcher
sees spaces in the model name and rejects valid commands like:
!model qwen3.7-plus --provider opencode-go
Skip block extraction for command messages while preserving it for
normal messages. Also preserve Slack thread_ts (top-level or nested in
message/container payload shapes) on native slash-command payloads so
session-scoped commands like /model apply to the intended thread instead
of a channel+user key the next threaded message never matches.
Surgical reapply of PR #43533 (originally against gateway/platforms/slack.py,
now plugins/platforms/slack/adapter.py). Thread-shape widening credit also
to #66310.
Follow-up to the #43755 salvage:
- Update the strict _Pool doubles in tests/run_agent/test_run_agent.py to
accept api_key_hint and assert it carries the agent's failed key.
- Add a real-CredentialPool regression (no mocks) proving the hint routes
exhaustion to the entry whose key actually failed, not pool.current(),
plus the no-hint baseline (#43747 wrong-entry marking).
recover_with_credential_pool() called mark_exhausted_and_rotate() without
api_key_hint, causing it to fall back to current() or _select_unlocked().
When a prior rotation left current() as None, _select_unlocked() returned
the NEXT (healthy) entry instead of the one that actually failed — marking
the wrong credential as exhausted (#43747).
Extract the current API key from agent.api_key (or pool.current().runtime_api_key
as fallback) and pass it as api_key_hint to all 4 call sites.
removeSession/archiveSession now pin their tombstone via beginSessionMutation
until the RPC settles (finally → endSessionMutation), keyed per id so
concurrent deletes across worktrees stay independent. This lets the
projects.tree prune hold the optimistic removal through the whole in-flight
window — no shared lock, no serialization.
Drops the now-redundant post-success re-filter band-aid in archiveSession:
the refresh honors tombstones generically, so there's nothing left to win.
A refresh racing an in-flight delete/archive resurrected the just-removed
row: refreshSessions repopulated $sessions straight from the backend page
ignoring tombstones, and the projects.tree prune dropped a tombstone the
moment its id left scoped_session_ids — so the grouped lane un-filtered it
too. The row only vanished on a later refresh once the RPC finally landed.
refreshSessions now filters tombstoned rows (and their lineage tip) before
merging. The prune keeps a tombstone while its mutation is still in flight
locally via $sessionMutationsInFlight, then hands it back to the normal
scoped-based prune once settled.
Clicking '+' on a collapsed worktree/branch lane (or repo) created a session in a folder the user couldn't see. It now force-expands the target node.
The root cause was that collapse state was stored as an XOR override of defaultOpen. defaultOpen flips for a worktree lane (collapsed while empty, open once it holds a session), so an explicit expand of an empty lane silently re-read as a 'collapse' the moment the lane gained its first row - collapsing the very lane you'd just opened to work in. Store the resolved open/collapsed boolean per node instead ($sidebarWorkspaceNodeOpen), which survives the default flip; one-time migration off the old set. The review file-tree shares this store and is updated to match.
The sidebar row and the pane tabs each painted their own dot from different data: the sidebar read color from $sessionColorById[id] (map-only, no resolver fallback) layered with live state, while a tab painted a flat 'accent' color via sessionColorFor (map + fallback). A session older than the recents page missed the map, so the same session showed grey in the sidebar and its project color in the tab.
Add a single SessionStatusDot primitive keyed by the stored session id (the key every live-state atom already uses) that resolves color (override -> project, with fallback) and live state (working/needs-input/stalled/unread/background) itself. The sidebar row, session tiles, and the main workspace tab all render it, so a session's status/color can never disagree across surfaces. Tabs gain the full live status (pulse etc.), not just a static color. Collapses the now-orphaned generic 'accent' tab-dot path into the one primitive.
Clear the coding rail while a session's workspace changes and discard late
Git status results for the previous cwd. This prevents Ctrl+Shift+B from
opening a worktree dialog against an old branch.
Use the trusted publisher's scoped GitHub token when resolving the pinned
gh-image release. This avoids GitHub-hosted runners exhausting their shared
anonymous API rate limit before the evidence publisher can run.
- ⌘W on the main tab promotes the next stacked session tab into main
- '+' / ⌘T open a new 'New session' tab, unlisted until first message
(no sidebar pollution); multiple new-session tabs allowed
- tile resolves its own row via by-id lookup once it has a message
- Extract resolveSessionColor(): the sessionColorFor fallback no longer
re-implements the override -> project-color precedence that the
$sessionColorById computed already owns.
- tileStoredRow: collapse the nested project-tree walk into a flatMap +
find, matching the local style.
The no-native-title guard (added on main after this PR's first CI run)
bans native title= on <button>. Wrap the shared ChoiceButton in the
themed <Tip>, which renders the child untouched when the label is falsy
so the live card is unaffected.
A tab group rendered only the active pane's content, so every tab switch
unmounted and remounted the whole surface — revisiting a session tab
re-measured and re-scrolled the thread from scratch, visibly shifting
layout each time.
Panes that have been active in a zone now stay mounted in absolutely
positioned layers; the inactive ones hide via visibility (keeping their
layout box, so scroll positions and measurements survive) with
pointer-events disabled and aria-hidden. Mounting stays lazy — a pane
first mounts when first activated — so a boot-restored tab stack still
doesn't resume every session up front, and panes that leave the zone
(closed / moved) unmount as before.
Opening a session in a new tab left the tab titled 'Session' with no
project-color dot until new activity landed the row in the paginated
recents list. Two gaps:
1. tileTitle/tileAccent/tileDragPayload only looked the stored row up in
$sessions (the recents page). Sessions opened from a project group are
often older than that page, so the lookup missed entirely. Resolve
through the project tree as a fallback (tileStoredRow) and re-sync pane
titles/accents when $projectTree loads.
2. Even with the row present, liveSessionProjectId returned null for a
session whose cwd sits outside its recorded git_repo_root (mid-session
relocation / sibling worktree), because the cwd-under-root guard ran
before the explicit-project folder match. The backend tree groups such
rows under the project; the client now agrees — an explicit folder
match is authoritative, only the auto-project (repo root) fallback
still needs cwd-under-root confidence.
sessionColorFor also computes directly (overrides -> project color) when
the row isn't in the $sessionColorById map, which is keyed over $sessions
only.
Keep the live PR review comment polling for a short grace period after
visible jobs complete. Preserve composite-action jobs in timing and live
status collection, and split the timing HTML from its linked review-status
artifact.
Builds on the skipped-clarify card so it holds up beyond the timeout case:
- Extract a shared `ChoiceButton` (letter badge + label + row chrome) used
by both the live pending card and the settled skip card. The two blocks
had drifted into duplicated markup; now they can't diverge.
- Fix the hint copy. An empty `user_response` is emitted for BOTH a
server-side timeout AND a manual Skip (tools/clarify_tool.py) — there is
no field on the result that tells them apart — so asserting "This question
timed out" was wrong half the time. Neutral wording ("This prompt is no
longer waiting…") is correct for either, and the recover-your-answer path
now also helps someone who mis-clicked Skip. Updated en/ja/zh/zh-hant.
No behavior change to the live prompt or the follow-up-message flow.
Co-authored-by: SHL0MS <SHL0MS@users.noreply.github.com>
When a clarify prompt times out, the settled card collapsed to just
'Skipped' — the options were unrecoverable (the args carry them, the
renderer dropped them) and there was no way to answer late.
The skipped card now:
- renders the original choices, letter-badged like the live card
- clicking one drafts a quoted follow-up ('Re: "<question>" — my
answer: <choice>') into the composer via the insert bus. Enter sends
it; if the agent is mid-turn it queues like any other prompt.
- a hint line explains the question timed out and what picking does
No retroactive resolution of the expired request: the tool already
returned empty and the turn moved on — injecting into past context
would break prompt-cache and role-alternation invariants, and
clarify.respond on an expired id hard-errors (#56558). The follow-up
message path needs no backend change and works against old backends.
Answered clarifies and free-text (no-choice) skips are unchanged.
Interim UX for #44845 (durable ID-addressable clarify decisions).
The bundled @vscode/codicons font has no `git-fork` glyph (only
`git-fork-private`), so the Branch menu item rendered blank. Use
`repo-forked`, the actual fork icon, to match the inline GitFork action.
The row/tab context-menu 'Branch from here' item used the git-branch
codicon while the inline message action uses a GitFork icon. Switch the
menu item to the git-fork codicon so branching looks consistent across
surfaces.
* fix(ci): publish inline E2E evidence
Upload bounded screenshot evidence from E2E, then publish validated images
from a trusted workflow_run job to commit-pinned branches in the evidence repo.
Wait briefly for the live CI review comment marker before publishing, so
GitHub's read-after-write delay cannot leave an orphaned evidence branch.
* fix(ci): isolate privileged credentials from PR jobs
Keep App private keys and Docker Hub credentials out of PR-controlled
workflows. Use protected environments for trusted publishing and a public
repository variable for the App client ID.
* fix(ci): attach E2E evidence with restricted bot session
Replace the App-backed evidence repository publisher with gh-image uploads
from a dedicated bot session in the gh-image environment.
* fix(ci): publish validated E2E evidence from forks
Let the trusted default-branch publisher handle bounded, validated evidence
artifacts from fork PR CI without checking out or executing fork code.
Branching/forking a chat replaced the primary chat (setActiveSessionId +
setSelectedStoredSessionId + navigate). Instead, open the branch as its own
session tile tab in the center zone and reveal it, leaving the parent chat
exactly where it is — mirroring openNewSessionTile. All branch entry points
(message GitFork button, /branch and /fork, sidebar Branch, tile Branch) go
through forkBranch, so this covers every surface.
Update the gateway's live turn projection after an accepted correction so a warm session resume does not add the stale original prompt beside the persisted correction.
Cover the inference-time correction path end to end and assert both redirect entry points refresh the live user text.
The base class spawns _keep_typing for every adapter — a 2s refresh loop
that runs for the whole turn, including while the stream consumer sends
the response — but RelayAdapter inherited BasePlatformAdapter's no-op
send_typing. The loop ran every turn and emitted nothing, so relay-fronted
chats (hosted Discord/Telegram/Signal/Slack) never showed 'is typing…'.
The rest of the pipe already existed on both sides: OutboundOp "typing"
is in the wire contract (contract_version 1), and every connector-side
sender implements it (Discord POST /channels/{id}/typing, Telegram
sendChatAction, Signal sendTyping, Slack assistant status).
This bridges the tick onto the existing outbound frame, mirroring send():
- _with_scope re-attaches the tenant discriminator (metadata.scope_id, or
user_id for DMs) — the connector's routedEgressGuard wraps ALL ops, so
an undiscriminated typing frame is declined like a bare send would be.
- the Phase 1.5 per-frame platform tag routes typing through the platform
the chat lives on for multi-platform gateways.
Best-effort by design: transport errors are swallowed (typing is cosmetic;
the next tick retries), no transport is a silent no-op, and stop_typing
stays the base no-op (platform indicators self-expire). Additive within
contract_version 1 — an older connector returns an unsupported-op result
we ignore. Contract doc §4 updated (typing now carries metadata?).
* test(desktop): cover correction resume without duplicate prompts
Exercise a live composer correction, switch away and back before the
response settles, and assert both user turns retain their order and occur
exactly once.
* test(desktop): cover correction warm resume during tool run
Exercise a correction accepted at a foreground-tool boundary, a switch through a persisted session, and the warm resume back. Assert the original prompt and correction remain singular and ordered.
Model settings was the only page that kept its shape while loading; the rest
flashed a centered spinner (LoadingState) or empty placeholders. Standardize on
skeletons that mirror the settings rhythm.
- Add shared SectionHeadingSkeleton / ListRowSkeleton / SettingsSkeleton to
settings/primitives.tsx (mirror SectionHeading + ListRow).
- Convert keys, providers, sessions, gateway, custom-endpoints, and config
(non-model) from LoadingState to SettingsSkeleton; remove now-dead LoadingState.
- billing: BillingSkeleton (summary cards + sections) on first load instead of
"—" placeholder cards.
- pet: skeleton grid on first load instead of a premature "unreachable" message.
- Iterate ROUTINE_COMPRESSION_STATUS_SAMPLES (formatted from the source
constants the emit sites use) through _prepare_gateway_status_message on
every chat platform — emission-wording drift now fails the suite without
re-copying literals.
- Extend the pinned NOISY_STATUS_MESSAGES with the buffered retry chatter
and post-#69332 wordings.
- New VISIBLE_COMPRESSION_MESSAGES negative suite: manual /compress
headlines and abort/failure notices must never be swallowed by the
widened regex.
Post-sweep audit of every compression status emission (conversation_loop,
turn_context, conversation_compression): the buffered overflow/attempt-cap
retry chatter (🗜️ 'Context too large…', 'Compressed X → Y, retrying…',
'Context reduced to…'), the #69332-reworded auto-lower notice
("Auto-lowered this session's threshold…"), the aux-provider-unavailable
notice, and the concurrent-compression skip all leaked past
_TELEGRAM_NOISY_STATUS_RE on chat platforms. Add anchored alternatives for
each; the ', retrying'/'— compressing' anchors keep manual /compress
feedback ('Compressed: 30 → 12 messages') and failure/abort notices
visible per the deliberate carve-outs.
Also extract every routine compression status string into importable
template constants in agent/conversation_compression.py (single source of
truth shared by all emission sites), so tests can iterate the actual
emitted wording instead of hand-copied literals.
Treat HERMES_DESKTOP_HERMES as an authoritative deployment override.
This keeps the Nix desktop package on its matching immutable Hermes CLI
instead of falling through to install.sh when a best-effort version probe
fails or times out.
Surface audit for #64438 consistency: cli.py and acp_adapter had their
gates removed by the salvaged #63630 commit (each with a behavioral
test). The remaining manual-compress surfaces never gated on
compression.enabled — pin that contract so a gate can't regress in:
- gateway/slash_commands.py _handle_compress_command (new behavioral
test: compression_enabled=False still compresses, force=True).
- tui_gateway/server.py — all three manual routes (session.compress
RPC, command.dispatch compress branch, slash.exec mirror) plus the
compute-host slash.compress/session.compress controls converge on
_compress_session_history; new test pins the helper ignores
agent.compression_enabled and passes force=True.
Main renamed the ACP slash command from /compact to /compress after #63630
was written; update the salvaged status line, comment, and behavioral test
to dispatch the command that actually exists.
compression.enabled: false is documented (agent/conversation_loop.py
overflow path) as disabling *automatic* compaction only — the terminal
context-overflow error explicitly tells users to run /compress manually,
and the gateway handler has never gated on the flag. But the classic CLI
(_manual_compress) and the ACP adapter (/compact) refused with
'Compression is disabled', leaving users at a full context with no
manual escape hatch on those surfaces.
Remove the stale gates (they predate the overflow-path design; the CLI
gate came from the original /compress commit's boilerplate) and unify
force=True across all manual-compaction call sites: ACP /compact and the
TUI's _compress_session_history (manual-only helper) now bypass the
summary-failure cooldown exactly like the CLI and gateway already did.
Also reword the ACP /context status line so a disabled-compression agent
no longer implies /compact is unavailable.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Integration test for the two compaction layers landing this sweep:
#63144's discovery-scope fix (archived rows surface from the current
session) and #69334's bookend bounding (summary exclusion + content
caps). A single compacted session exercises both: the archived FTS hit
must surface while the compaction handoff at the session tail stays out
of bookend_end and long messages stay capped.
Addresses @teknium1's second review round on #63144:
1. _is_compacted_message now checks both active=0 AND compacted=1.
Previously checked active=0 alone, which also matched rewind/undo rows
(active=0, compacted=0) that must stay hidden.
2. Added _is_compression_ended() — checks only the session's own
end_reason, not the lineage-wide has_compression_hop flag. This prevents
delegation children living under a compression continuation from leaking
through the lineage filter.
3. _discover lineage skip now uses is_ended_session (session-level check)
instead of has_compression (lineage-level flag).
New tests:
- TestRewindExclusion: rewind rows stay hidden alongside compacted rows
- TestCompressionEndedHelper: session-level end_reason checks
- TestLegacyContinuationPlusDelegation: delegate child excluded while
compression ancestors surface
78 passed (71 + 7 new).
After context compaction, pre-compaction content was invisible to
session_search — a memory black hole. The _discover() skip logic
filtered both same-session and same-lineage hits unconditionally,
without distinguishing compression-summarised content (gone from live
context) from delegation children (still visible to the parent agent).
Reworked _resolve_to_parent to return (root_id, has_compression_hop),
checking end_reason='compression' on every hop during the same
db.get_session() traversal — zero extra queries.
_discover() now has three compression-aware paths:
- In-place compaction: FTS hits on active=0 (compacted=1) rows pass
through even when raw_sid == current_session_id
- Legacy rotation: lineage hits pass through when has_compression_hop
is true on either side of the chain
- Delegation children: still excluded (no compression edge)
18 new tests covering all three scenarios + unit tests for the helpers.
Addresses Teknium's review feedback on #6256.
Closes#13840, #13841.
With compression.in_place defaulting True the cron session id never
rotates; get_compression_tip returns the input id. Pin that the
salvaged #67188 fix titles/ends the ORIGINAL cron session in that path
(and for falsy tip returns), i.e. zero behavior change when no
compression rotation happened. Also maps colingreig's contributor
email for attribution CI.
Hardening for the #59114 salvage: generate real standalone and merged
handoffs with the current compressor and assert classify_summary_content
agrees with _is_context_summary_content and is_compaction_summary_message
on every emitted shape (behavior contract, not a format snapshot), incl.
the flag-stripped DB-reload copy. Also adds the contributor email mapping
for @israellot.
A context-compaction handoff is persisted as an ordinary history message
but is not a real turn. The ACP history replay streamed it as a bare
user/agent message chunk, dropping the in-process _compressed_summary
marker, so ACP frontends (editors, vscode-hermes) rendered the entire
handoff as a regular message.
Tag replayed summary chunks under _meta.hermes (ACP's extensibility
channel), covering all three persistence shapes the compressor emits:
- standalone role="user" handoff -> compactionSummary: true
- standalone role="assistant" handoff (alternation-driven role pick)
-> compactionSummary: true
- merge-into-tail message (preserved tail content + appended summary)
-> containsCompactionSummary: true, a distinct key so clients that
collapse standalone summaries cannot hide the preserved real content
Detection honors the in-process metadata flag and falls back to a new
ContextCompressor.classify_summary_content() content classifier
(standalone/merged/None), so it also works for a DB-reloaded session
that lost the in-memory flag. _is_context_summary_content is now a thin
wrapper over the classifier, keeping existing callers unchanged.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Add components/ui/progress.tsx — one rounded track + animated fill that owns
role="progressbar" + aria. Migrate the hand-rolled bars (updates overlay,
onboarding, install overlay, billing usage, pet hatch) onto it.
- Pet's bar was never a color variant (--primary and --ui-accent are the same
brand color); its only real difference is the sliding indeterminate, now an
`animated` prop. Color stays a normal `fillClassName` override.
- Billing usage now uses the plain primitive — dropped the bespoke dither
track, inset shadow, and danger nub; tone rides `destructive`/`fillClassName`.
- Drop the redundant emoji "no saved card" row description; the page-level
warn banner is the single explainer.
- Rename pet CSS .pet-progress* -> generic .progress-slide.
* fix(desktop): show steer glyph in busy composer
Restore the steering-wheel glyph when a typed draft redirects a live turn.
Add a real Electron E2E regression test for stop, steer, attachment queueing,
and a paused queue after Stop.
* feat(desktop): add queue action beside steer
Restore an explicit queue action in the former steering slot while a typed
correction is ready. Keep the primary steering-wheel action for redirecting
the active turn.
* fix(desktop): retain dictation beside queued drafts
Keep the dictation microphone visible when a typed busy draft exposes the
separate queue action. Cover the combined controls in the real Electron flow.
* fix(desktop): place queue beside the busy action
Keep the Queue message control after the read-aloud toggle and directly before
the shared Stop/Steer action. Cover the rendered control order in Electron E2E.
* ci: surface E2E screenshots in review comment
* ci: mark completed review commits in past tense
* ci: surface approved sensitive-file reviews
* ci: link sensitive files to reviewed changes
* ci: stage desktop E2E visual evidence
Track screenshots newly introduced against main and package visual diffs for a trusted publisher.
* fix(ci): pass E2E evidence output paths
Supply the manifest and staging-directory arguments required by the screenshot status helper.
* fix(ci): download the OSV SARIF artifact
Match the artifact name and result filename emitted by the pinned upstream reusable workflow.
* fix(desktop): refresh composer branch after worktree creation
Route new worktree sessions through the shared workspace target handoff so
composer git status follows the created worktree instead of remaining on the
main checkout. Add an Electron E2E regression test for Ctrl+Shift+B.
* fix(desktop): fixme flaky test
The theme engine picked light/dark adaptation from the HOST terminal
(detectLightMode) even when the skin authors its own background — which
the TUI then paints onto the terminal via OSC-11. On a light-mode Apple
Terminal without truecolor, a pure-black skin (e.g. Bloomberg) got its
foregrounds ansi256-bucketed *for a light background that no longer
exists*: theme.color.text became 'ansi256(214)', which also fails the
OSC-10 hex gate, so the terminal default fg stayed the light profile's
near-black — markdown body text rendered black-on-black.
fromSkin now resolves polarity from the skin's authored background when
present (skinIsLight), uses that background as the reference canvas for
the derived tone ladder and contrast adaptation, and only falls back to
host detection for skins without a background (they render on the
terminal's own surface). The paired light_colors/dark_colors pick in
themeForSkin follows the same rule.
Pin a provider-agnostic "Out of credits" panel after a billing-classified turn so
the one recovery action (Nous → /topup, other providers → their billing page)
stays visible instead of scrolling away as prose.
Open the shared ConfirmPrompt (full-width, themed, same lane as approval/clarify)
rather than a truncating status-bar notice. One recovery action: Nous → /topup
(opens the rich billing overlay), other providers → their billing page, or /model
to switch when there's no URL. The transcript keeps the full guidance; the dialog
is the concise actionable layer.
On a billing wall, raise a sticky, billing-specific toast (never the generic
error toast) and a persistent in-composer banner for the active session — both
with one recovery action: Nous → in-app Settings → Billing, other providers →
their billing page (deep-linked). The banner reuses the shared StatusRow chrome
(no bordered alert, Codicon glyph, shared buttons), and the composer stays usable
so slash commands keep working.
Detect a billing wall once (agent/error_classifier → FailoverReason.billing) and
map it to a recovery link + label in one place, then carry that structured
BillingBlock to every surface instead of re-parsing free-form error text per
surface.
- agent/billing_links.py: provider-agnostic slug/host → (label, billing URL)
table (single source of truth), Nous-aware (is_nous routes to the in-app flow);
unknown providers degrade to a readable label with no invented URL.
- conversation_loop: both billing exit paths return a billing_block through one
helper; the guidance message carries the derived URL for every provider.
- gateway forwards billing_block on message.complete (it was dropped).
- @hermes/shared: BillingBlock type shared by desktop + TUI.
* test(desktop): e2e test for interim assistant message preservation (#65919)
Adds a Playwright E2E test that reproduces the fix from PR #65919 across
all three layers (agent core → tui_gateway → desktop renderer). The mock
inference server is upgraded with a multi-turn scripted response that
exercises several interleaved patterns:
1. text + tool_call → should produce an interim message
2. text + tool_call → another interim message
3. no text + tool_call → NO interim (no visible text alongside tools)
4. text + tool_call → another interim message
5. final answer (stop) → message.complete, different from all interims
Two describe blocks exercise display.interim_assistant_messages both on
(default) and off:
- ON: all interim texts + the final answer visible in the transcript
- OFF: only the final answer visible, all interim texts wiped
Also fixes a footgun: test:e2e now runs `npm run build` as a pretest
hook so the renderer dist/ is always fresh. Previously, running
`npx playwright test` locally would silently load a stale dist/ that
predated renderer fixes — the python backend ran from source (had the
fix) but the renderer was frozen in an old bundle. CI already built
fresh, so the explicit build step there is removed to avoid duplication.
* test(desktop): e2e sidebar states — background dot, subagent, cross-session
Add sidebar-states.spec.ts with three E2E tests exercising the desktop
sidebar's session dot states driven by real gateway events:
1. Background process dot appears during a terminal(background=true)
call and disappears after auto-dismiss; subagent (delegate_task)
runs concurrently; final answer is visible in the transcript.
2. Background dot remains visible while a subagent runs concurrently
(longer sleep 5 background process so the dot is catchable).
3. Cross-session dot transition: start a turn with a background process,
wait for the turn to complete, open a new session, then verify the
original session's dot transitions from 'background running' to
'finished — unread' when the background process exits.
The mock server gains SIDEBAR_SCRIPT and SIDEBAR_CROSS_SCRIPT trigger
keywords that return tool_calls for terminal(background=true) and
delegate_task — the agent executes these for real (real background
process, real subagent), so the tests assert against genuine gateway
events rather than mocked UI state.
Verified: 3 passed (1.2m) under cage headless wlroots.
* test(desktop): e2e tests for tile-unread bug (tab passes, split fails)
Two scenarios for the tile-unread bug where a session that finishes
while visible on-screen gets the green 'finished unread' dot even
though the user is looking right at it.
The unread check in handleTransition (session-states.ts:174) only
compares against $selectedStoredSessionId and ignores $sessionTiles,
so a session visible in a tile gets marked unread even though it's
on screen.
1. TAB (hidden, PASSES): ⌃-click opens the session as a stacked tab
that is NOT visible on screen. The unread dot IS correct here —
the user isn't looking at it.
2. SPLIT (visible, FAILS): drag the session row to the workspace's
right edge to create a side-by-side split tile. Both sessions are
visible on screen. The unread dot is WRONG — the session is visible
in the split tile, so it should not be marked 'unread'. This test
is RED until the fix lands.
Also adds explicit page.screenshot() calls at key assertion points in
sidebar-states.spec.ts so the trace viewer has full-res captures of the
sidebar dot states during the test.
* test(desktop): cover compression and queued stop lifecycle
Add real desktop E2E coverage for session compression continuation and
queue parking after an explicit Stop. Extend the mock server with a
blocking scripted turn and submitted-prompt assertions.
* test(desktop): cover busy composer submit routing
Replace the invalid queued-stop E2E scenario: plain text redirects a busy
turn rather than entering the queue. Add focused submit-routing coverage for
plain text, slash commands, attachments, explicit Stop, and idle submission.
markVoicePlaybackInterrupted() / takeVoicePlaybackInterrupted() mirror
the backend latch in the renderer (the barge happens client-side, where
the audio plays). VAD barges and typing over playback mark it; the next
prompt.submit carries interrupted:true, which the TUI gateway latches
into the model note.
CLI: the VAD monitor's playback cut and the record-key interrupt both
mark the latch; the chat path prepends the note via the existing
_prepend_note_to_message channel. TUI: _tts_stream_stop() grows a
user_barge flag (False for /voice off — a mode change isn't an
interruption; no-op when speech already finished), the VAD monitor
marks on cut, prompt.submit accepts interrupted:true from clients, and
_run_prompt_submit pops the latch into the run message.
mark_speech_interrupted() / take_speech_interrupted(): a one-shot,
TTL'd (120s) latch plus SPEECH_INTERRUPTED_NOTE. Barge-in paths mark
it when they cut live speech; the next turn's submit path pops it and
prepends the note to the model-bound message — API-call local, never
persisted, so history and prompt caching are untouched.
/api/audio/speak-stream WebSocket: one socket + one Web Audio clock per
reply. The renderer feeds raw LLM deltas as they arrive; the server
cuts sentences with the shared chunker and streams int16 PCM back while
generation continues — speech overlaps generation with no per-sentence
connection or synthesis gaps. Falls back to the POST data-URL path for
old backends / non-chunked providers.
Barge-in runs a MediaRecorder on the monitor's stream the whole time
playback is live (rotated while quiet to bound pre-roll); talking over
the agent cuts playback and the complete utterance goes straight to
transcription and submit. /api/audio/transcribe returns 200/"" for
no-speech results so quiet turns re-listen instead of toasting a 400.
Turn deltas feed a per-turn TTS pipeline; the post-complete speak_text
call survives only as a fallback. session.interrupt, /voice toggles,
and new turns cut in-flight speech. VAD barge-in emits
voice.interrupted at detection, then the captured interruption goes out
as voice.transcript — the same event the TUI already submits as a
spoken turn.
The streaming gate broadens from ElevenLabs-only to any provider that
passes check_tts_requirements(). In continuous voice mode a mic monitor
runs during playback: talking over the agent cuts TTS at detection
while the monitor keeps recording, then the captured interruption is
transcribed and queued as the next turn (process_loop's auto-restart
stands down while the capture owns the mic). New voice.barge_in config
key, default true.
listen_for_speech(): sustained-RMS speech detection with the noise
floor calibrated against audible playback (speaker bleed doesn't
self-trigger). With capture=True it keeps a rolling pre-roll buffer and
records through to a silence endpoint, so the interruption is
transcribed from its first syllable — detection alone loses the opening
words to the detector's sustain window plus the mic re-open.
transcribe_recording() maps no_speech provider results to a successful
empty transcript (silence, not an error).
An STT provider that hears no words is reporting silence, not failing.
ElevenLabs/xAI empty-transcript errors now carry no_speech so live
voice loops can re-listen quietly instead of surfacing an error on
every pause.
stream_tts_to_speaker() drops its hardcoded ElevenLabs client for
resolve_streaming_provider() + SentenceChunker, so any provider speaks
sentence-by-sentence while the model is still generating. Markdown
stripping also drops emoji — providers stall on them or read them out
loud.
StreamingTTSProvider ABC + registry + resolve_streaming_provider() —
ElevenLabs (pcm_24000) and OpenAI (response_format=pcm) stream chunked
PCM; every other provider keeps its configured voice and falls back to
per-sentence sync synthesis. SentenceChunker is the one incremental
cutter every surface shares: sentence-boundary cuts on the delta
stream, <think> blocks stripped even when split across deltas, short
fragments merged forward so they never stall as tiny clips.
Keep a live session projection from adding its user turn when the latest
persisted row already represents that same inflight prompt. Add real Electron
coverage for fast and cold resume with idle and background-inference sessions.
Post-merge review polish of #68722 + #68761:
- Usage: drop the manual refresh + 'Updated Xm ago'; billing queries now
refetchInterval-poll while the page is mounted, like every other data view.
- No card: lead the page with a warn notice naming the blocker + 'Add card ↗',
so the silently-disabled buy/auto-refill controls have an obvious cause.
- Layout: unify on a shared SettingsCard/SettingsSection primitive; collapse the
three floating one-row sections (Payment / One-time top-up / Automatic refill)
into a single divide-y 'Payment & credits' card.
- Dev fixture switcher: relabel as a wrench + 'preview' dashed control so it
reads as the DEV-only tool it is (compiled out of production).
Removed formatUsageUpdatedAgo/oldestUpdatedAt/UsageRefreshRow + their tests;
added no-card-notice and no-manual-refresh tests. vitest 107 green, tsc + eslint clean.
* test(desktop): e2e for warm-route resume render jitter
Pre-seeds a 32-message session into state.db, boots the app, does a
cold resume (populates warm cache), navigates away, then clicks back
(warm resume). A MutationObserver + innerHTML-length poll detects
whether the transcript is re-rendered after the initial warm-cache
paint — the jitter bug where syncSessionStateToView fires twice
(warm cache paint, then session.activate RPC reconcile).
* fix(desktop): warm-route resume jitter from double setMessages
The warm resume path in resumeSession() calls syncSessionStateToView
twice: once for the warm cache paint, then again after the
session.activate RPC reconciles messages. The second call created new
message objects via toChatMessages (different references, same content),
and flushPendingViewState's sameMessageList guard used reference
equality per slot — so it always failed and setMessages fired a second
time, causing a visible transcript re-render.
Replace sameMessageList (reference equality) with chatMessageArraysEquivalent
(deep content comparison: id, role, parts, pending, error, etc.). This
was already used by the cold path's fast-path guard; the flush guard
was the last holdout using shallow reference equality.
Also updates the e2e test to use textContent polling on the first
message element (instead of innerHTML on the full viewport) to avoid
false positives from metadata-only DOM changes.
* test(desktop): e2e for warm resume after background inference
Extract the render counter (MutationObserver + text-content poll) and
assertion into reusable helpers. Add a second test that sends a message,
waits for the mock response to complete, navigates away, then warm-
resumes — verifying the warm cache already has the completed turn from
message.complete events and no second paint occurs.
The input fix's sibling, hit immediately after: the composer was themed but
AGENT text went black-on-black the same way. Root cause is the class, not
the call site — markdown body, borders, and every token rendered without an
explicit color falls back to the terminal's DEFAULT foreground, which
belongs to the HOST profile's polarity, not the skin's. A dark skin on a
light terminal repaints the backdrop via OSC-11 while thousands of
default-fg cells stay near-black.
Chasing every <Text> is unwinnable. Instead own the default itself: when a
skin authors a background (the existing opt-in), paint the default
foreground from the resolved theme's text color via OSC-10. Every unthemed
token — present and future — re-bases onto the skin atomically, exactly
like the background.
terminalModes: the OSC-11 slot generalizes to defaultColorSlot(10|11) —
same paint/clear/exit-restore contract, tracked per slot, so a skinless
session still never touches the terminal. reapplyTheme repaints the fg too:
polarity flips swap paired palettes, moving the text tone while the
background stays.
Tests: slot contract runs table-driven over both OSC codes; handler test
pins the invariant (default fg == theme text; dropping the background
releases both defaults). Suite 1344✓, typecheck/lint/prettier clean.
Live-repaint's composer gap: flip a light terminal to a dark skin and the
input goes black-on-black. The placeholder was already explicit truecolor
(theme muted), but TYPED text rendered with no color at all — the terminal's
default foreground — in both paint paths:
- the Ink render (<Text wrap="wrap">{rendered}</Text>, no color), and
- the fast-echo bypass, which writes raw cells straight to stdout.
The skin owns the background (OSC-11) but the default fg still belongs to
the host terminal's polarity, so any skin/terminal polarity mismatch made
input invisible. Every other transcript line already paints
theme.color.text (the completed inputBuf rows directly above the composer).
Give TextInput a color prop and paint both paths with it: the Ink <Text>
(chalk re-opens the outer color after the placeholder chips' embedded [39m
closes; INV cursor/selection cells never touch fg) and the fast-echo write
via colorizeEcho — same explicit-truecolor-only rule as colorizeHint, so
the bypass cell can't flash terminal-default before the next frame. All six
TextInput sites (composer, prompts, masked, billing ×2, session switcher)
pass theme text; no color ⇒ passthrough, unthemed inputs keep the terminal
default.
Tests: colorizeEcho SGR wrap + passthrough contracts; full ui-tui suite
1338✓; typecheck clean.
Removes Homebrew and PyPI wheel/sdist as Hermes distribution paths while
preserving the supported source, Docker, and Nix workflows.
Changes:
- Removes the Homebrew formula, PyPI publish workflow, sdist manifest
(MANIFEST.in), and wheel/sdist release-attachment logic from scripts/release.py.
- Keeps setuptools metadata and entry points required by editable installs
and Docker/Nix builds, but adds a setup.py guard that rejects wheel/sdist
builds outside a sealed Nix derivation (HERMES_NIX_BUILD=1).
- Removes pip/Homebrew install detection, PyPI update checks, the pip
self-update path, the deprecation-banner state, the postinstall subcommand,
wheel data-directory fallbacks in agent/i18n.py and hermes_constants.py,
and the ACP Registry manifest/version-lockstep release logic.
- Adds /nix/store/ path detection so `nix run` / `nix profile install`
installs (which don't set HERMES_MANAGED) are correctly identified as
"nix" rather than falling through to "git"/"unknown".
- Retired install-method values ("pip", "homebrew") in existing
.install_method stamps (both code-scoped and home-scoped) are ignored by
the allowlist reader and fall through to "unknown" instead of resurrecting
a retired enum value.
- Updates Nix packaging to ship bare runtime data (locales, optional-mcps)
through store symlinks and wrapper env vars instead of wheel data-files.
- Removes the ACP Registry manifest/icon and their version-lockstep tests.
- Deletes or rewrites packaging, pip-update, Homebrew, and ACP Registry
tests; adds parametrized coverage for the packaging build guard covering
BOTH sdist and wheel paths (the guards live in separate cmdclass entries
— a passing sdist test proves nothing about the wheel path).
- Updates installation/platform documentation and related user-facing copy.
- Adjusts the supply-chain scan so deleted install-hook files do not trigger
a finding, while additions or modifications still require the existing
ci-reviewed label gate.
Supported installation paths (unchanged):
- git installer (install.sh)
- Docker
- Nix/NixOS
- editable development installs (uv sync, uv pip install -e ., pip install -e .)
Exercise the full Electron, gateway, and mock-provider submit path while
same-chat route query tokens churn during session creation. Assert the mock
provider receives the prompt and its streamed response reaches the transcript.
* fix(desktop): route /compress through session.compress RPC with transcript replacement
Salvages #44462, #53755, and #68218 into a single canonical fix for the
desktop /compress cluster.
The desktop routed /compress through slash.exec, which sends it to the
_SlashWorker subprocess. Compressing a large session outlives both the
desktop's 30s WS timeout and the worker's 45s pipe timeout — the client
gives up, runExec's blanket catch swallows the error, and command.dispatch
surfaces a misleading "not a quick/plugin/skill command: compress" (#44456).
Even when compression succeeded via the _mirror_slash_side_effects path,
the desktop never received the post-compress message list, so summarized
bubbles stayed on screen forever — /compress looked like a no-op.
This change routes /compress to the dedicated session.compress RPC (the TUI's
path), combining the best of all three PRs:
- 120s client timeout matching the TUI's HERMES_TUI_RPC_TIMEOUT_MS (#44462)
- Transcript replacement from the response `messages` via toChatMessages,
the same converter session.resume uses (#68218, teknium1 review on #44462)
- Session-isolation guard: updateSessionState only publishes for the active
runtime, so a late result after a session switch can't clobber the
foreground transcript (#53755, teknium1 review on #53755)
- Coalescing: dedup concurrent compress requests per session (#53755)
- Progress toast ("compressing context...") outside the transcript (#53755)
- Error unmasking in runExec: when slash.exec fails and command.dispatch only
adds "not a quick/plugin/skill command" routing noise, surface the original
worker error instead (#44462)
- /compact alias + focus_topic forwarding
Co-authored-by: AlliDev <AIalliAI@users.noreply.github.com>
Co-authored-by: PinkEVO <PINKIIILQWQ@users.noreply.github.com>
* feat(desktop): route slash commands with dedicated RPCs to those RPCs
Salvages #63513 — introduces a new `rpc` kind on DesktopCommandSurface so
commands with a first-class gateway @method handler bypass slash.exec /
command.dispatch entirely, and a `renderRpcResult` utility that shapes
each RPC's structured reply into readable transcript text.
Migrates 6 commands from exec() to rpc(...):
/agents → agents.list
/save → session.save
/status → session.status
/steer → session.steer
/stop → process.stop
/usage → session.usage
/compress stays as action('compress') — it needs transcript replacement
from the response `messages`, which the generic rpc path can't do (per
teknium1 review on #44462/#63513).
Also includes the json-rpc-gateway timeout message improvement: the error
now includes the configured timeout duration ("request timed out after 120s:
session.compress") so a user can tell whether the default 30s fired or a
per-call override.
Co-authored-by: Jelvin <SmallNew2003@users.noreply.github.com>
* fix(desktop): preserve provider choice during config initialization
* fix(desktop): preserve slash command and host compression semantics
Keep commands whose CLI behavior exceeds their current RPC contracts on slash.exec.
Propagate the full compression timeout through compute-host control, return structured
host compression outcomes with metadata, and retain successful compression feedback
in the desktop transcript.
Add regressions for timeout forwarding, host aborts and metadata sync, structured host
control responses, command routing parity, and numeric stop counts.
* fix(desktop): harden compression state handling
Preserve the invoking stored-session binding for delayed compression results,
normalize replacement histories, and serialize provider selection. Stabilize
gateway platform tests and guard the desktop Git facade during renderer teardown.
---------
Co-authored-by: AlliDev <AIalliAI@users.noreply.github.com>
Co-authored-by: PinkEVO <PINKIIILQWQ@users.noreply.github.com>
Co-authored-by: Jelvin <SmallNew2003@users.noreply.github.com>
The submit "session context drift" guard (regression 7acaff5ef / #54527,
partially fixed by 8c288760d and da52ffea1) aborted a prompt submission
whenever the selected stored id OR the route token changed mid-submit. Both
signals churn programmatically on a busy gateway, so on machines with
background streaming sessions, per-minute cron sessions, the Telegram surface,
or gateway-profile switches, essentially every send from a second chat aborted
silently: the optimistic message was dropped, the draft was left in the
composer, no error was shown, and prompt.submit never fired.
The false-positive churn sources were:
- selection null-resets — gateway-switch's setSelectedStoredSessionId(null)
on a gateway/profile switch or reconnect read as a switch away;
- search/hash-only route-token changes — overlays and side panels park state
in location.search/hash, so the pathname (the only part that selects a
chat) was unchanged yet the raw token differed;
- background-event active-ref retargets — createBackendSessionForSend's
3-prong check also watched activeSessionIdRef, which gateway events retarget
while other sessions stream (#47709 class), during a seconds-long
session.create round-trip.
New shared helper session-context-drift.ts reduces a route token to the chat it
targets (pathname only; the new-chat route is '__new__', non-chat routes null)
and reports drift only when selection or the routed chat moves to a DIFFERENT,
non-null chat that is not the submit's own target. Selection null-resets,
search/hash-only churn, and moves onto the submit target are no longer drift;
genuine user switches (click another chat, click New Session mid-submit) still
abort. Site A (submit.ts) routes all five guard points through the helper and
logs '[submit-drift-abort]' with a per-site phase; the post-create active-ref
check and baseline re-pin from 8c288760d are kept intact. Site B
(createBackendSessionForSend) drops the active-ref prong entirely — every real
switch retargets selection and route synchronously — and logs before closing
the orphaned session.
(cherry picked from commit b390e3a22ee0dacc94283cda1748e3bafae15313)
Co-authored-by: Kennedy Umege <kenmege@yahoo.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Live theme authoring's core loop — Hermes recolors the skin file it just
activated — repainted the TUI but not the GUI. The event path was fine
(post-#69533 the WS broadcast lands and ingestBackendSkin refreshes the
$backendThemes registry); the same-name apply guard also no-ops correctly
(it's what protects a manual desktop theme pick). The repaint was supposed
to come from the registry: the active theme IS that skin, its palette just
changed. But ThemeProvider memoized deriveTheme on [themeName, resolvedMode]
only, while deriveTheme reads the registry non-reactively via resolveTheme —
so the store update re-rendered the provider and handed back the stale
palette. Name switches repainted (themeName moves); recolors never did.
Add the theme stores (user/backend/registry) to the memo's deps — they are
deriveTheme's actual reactivity, same as the availableThemes memo directly
above. applyTheme is idempotent, and $backendThemes only publishes on a
real palette change, so no spurious repaints.
Tests: render ThemeProvider for real — activation applies; a same-name
recolor repaints (fails without the fix); an inactive-skin seed doesn't
touch the painted theme.
_emit and _broadcast_global_event were each building the JSON-RPC event
envelope — extract _event_frame and use it from both. Type the registry as
set[Transport] (protocol already imported), and cut comment bloat at the
call sites. No behavior change; suites stay green.
* chore(gitignore): ignore installer .install_method stamp
Salvage of #54855 by @drissman — rebased onto current main with root-scoped
rule and sister-marker comments alongside .update-incomplete.
Closes#66189
Root cause: scripts/install.sh writes <install>/.install_method but git
did not ignore it, so managed checkouts show ?? .install_method and
hermes update may autostash the untracked marker.
Fix: add /.install_method to .gitignore (repo-root only).
Verification: git check-ignore -v .install_method
* test(update): assert .install_method survives update autostash (#66189)
Add hermetic regression mirroring the .hermes-bootstrap-complete test:
adopt the real .gitignore, drop the installer .install_method stamp, run
the exact 'git stash push --include-untracked' the updater uses, and assert
the marker is neither swept nor reported dirty. Requested by hermes-sweeper
review on #67364.
Real-world failure from dogfooding the live-theme flow: display.skin was
already 'synthwave' in config, but the desktop never visibly applied it (the
activation event predated the WS transport fix / the connect). The desktop's
gateway.ready seed records the baseline WITHOUT painting (by design — never
stomp the persisted desktop theme on connect), so it believed it was synced.
Re-running 'hermes config set display.skin synthwave' then did nothing twice
over: the watcher signature (name, skin-file mtime) hadn't moved, so no
skin.changed fired; and even on an event, the desktop's name-equality guard
blocked the apply against the seeded baseline.
Two halves:
- hermes_cli: setting display.skin touches the named skin file so the
watcher signature always moves on an explicit set — a same-name re-affirm
now broadcasts skin.changed like any real move. Built-ins (no file) are
unaffected; a name switch already moves their signature.
- desktop: track whether the synced baseline was actually APPLIED vs merely
seeded at connect. A skin.changed matching a seed-only baseline is an
intentional apply and repaints; once applied, repeat same-name events stay
no-ops (protects a manual desktop-side theme switch from snap-back, incl.
across a reconnect re-seed).
When Codex returns 429 usage_limit_reached, Hermes persists the provider's
reset_at on the pool entry and freezes the credential until it elapses --
which can be days out for weekly windows. But the upstream window can
reopen EARLY: the user redeems a banked rate-limit reset (Codex CLI /
ChatGPT UI), upgrades their plan, or OpenAI resets the window. Hermes
never re-checked, so it kept erroring with 'Codex provider quota
exhausted (429); retry after Ns' until a manual re-auth rewrote the
tokens (issue #43747, externally-reset variant).
- hermes_cli/auth.py: add _probe_codex_quota_restored() -- a throttled
(5 min/token) GET of the Codex /usage endpoint; quota counts as
restored when every reported window is <100% used. Add
clear_codex_pool_quota_cooldowns() to lift 429/quota-shaped cooldowns
from persisted pool entries (DEAD and auth-shaped entries untouched).
- resolve_codex_runtime_credentials(): before surfacing a pool-only
cooldown as 'quota exhausted', probe upstream; on a positive probe
clear the cooldown and return the pool credential.
- agent/credential_pool.py: _available_entries() probes frozen
openai-codex entries (clear_expired path only) and unfreezes them when
upstream confirms the reset.
- agent/account_usage.py: a successful /usage reset redemption now
clears persisted pool cooldowns immediately.
Negative paths preserved: probe 429/exhausted/indeterminate keeps the
cooldown; read-only enumeration never probes; non-JWT tokens never
probe (no network in hermetic tests).
The cross-surface theme SDK's live-repaint relies on a gateway skin watcher
that polls config and emits skin.changed on any move. But that emit is
session-less and fires from a background thread, so write_json fell through
its (session-transport -> contextvar -> stdio) ladder to the module stdio
transport — which only reaches the stdio TUI (tee'd to the dashboard WS
publisher). WS clients (the desktop app, dashboard chat) never got it, so
'Hermes themes itself' repainted the CLI/TUI but not the GUI.
Add a live-transport registry (one entry per connected WS peer, maintained by
handle_ws) and a _broadcast_global_event primitive that fans session-less
announcements out to every connected client, falling back to write_json when
none are registered (stdio path unchanged). Route both skin.changed emits
(watcher + the /skin RPC) through it, so a skin switch from any surface
repaints all of them.
Backend-only; desktop already handles skin.changed and does not drop
session-less events.
Two narrow timing windows (reported by null-runner) silently downgraded a
mid-turn correction to a plain next-turn message on the desktop client:
- Turn-build window: a fresh turn flips running=True and builds the agent
asynchronously, so session["agent"] is briefly None. session.redirect
answered 4010 "unsupported", which the renderer's catch swallowed into a
lost follow-up. Queue the correction server-side instead and return
status="queued" — lossless, and honest about what happened.
- Stale runtime id after reconnect: session.redirect 404s on a sid the
gateway no longer maps. redirectPrompt now resumes the stored session and
retries once, mirroring stopPrompt, so a correction fired right after a
reconnect isn't dropped.
The desktop treats "queued" like "redirected": the correction reaches the
model either way, so it's recorded once as a real user message.
The zero-result fallback prefers messages_fts_cjk when built: exact
ranked token match for Latin runs unicode61 fused onto CJK, including
<3-char tokens the trigram leg can't recover.
A pure-Latin query (no CJK characters) routes to the unicode61
`messages_fts` table, whose tokenizer does not insert a boundary between
Latin letters and adjacent CJK characters. Content like "修改youer服务端" is
indexed as a single token, so `search_messages("youer")` returned zero
results even though the substring is present, and the Latin path had no
fallback.
Add a zero-result trigram fallback to the pure-Latin path: when the
unicode61 search misses, retry against the existing `messages_fts_trigram`
table, which matches substrings regardless of word boundaries. The fallback
is gated on `_trigram_available` and on every token being >=3 chars (the
trigram minimum), and only fires on a zero-result miss, so successful Latin
searches keep their unicode61 ranking unchanged.
The trigram query construction shared with the CJK path is extracted into a
`_run_trigram_search()` helper; the CJK branch is refactored to use it with
no behavior change.
Adds regression tests in tests/test_hermes_state.py::TestCJKSearchFallback.
A stray tsc run can emit foo.js next to foo.ts under apps/shared/src or
apps/desktop/src. .gitignore hides the artifact from git status, but Vite
resolves extensionless imports .js-before-.ts, so the renderer silently runs
the stale compiled copy.
tsc -b . --clean already knows the emit graph and deletes
matching outputs. Run it before vite in all dev scripts.
This bit for real: a Jul 16 artifact of websocket-url.js predated the #68250
getGatewayWsUrl contract change ({ ok, wsUrl } IPC result), so its old
'if (fresh) return fresh' handed the whole result object to new WebSocket(),
dialing ws://127.0.0.1:5174/[object%20Object] on every boot. The desktop app
could never connect, and the failure survived reboots and cache wipes because
the poison lived in src/.
JsonRpcGatewayClient.connect() now rejects non-ws:// URLs with a readable
error instead of letting new WebSocket() coerce an object into
[object%20Object], so any future contract skew fails diagnosably.
- Use the fork glyph for branch and a sine wave for read aloud (all one lib now)
- Extract compact "2h ago" into formatAgo() in lib/time.ts (+ ageDays locale string)
- Cover formatAgo with a unit test
Extract the open_preview emitter into a shared tools/desktop_ui bridge
(one gateway-injected sink, routed by HERMES_UI_SESSION_ID) and add a
second desktop-gated tool on top of it:
- focus_pane(chat|files|terminal|review|sessions) -> pane.reveal event.
The desktop runs each pane's own reveal path (revealDesktopPane table)
and only acts on the active window -- a background turn never moves the
user's focus (desktop AGENTS.md: offer, don't hijack).
open_preview now emits through the same bridge. Both tools are check_fn
on HERMES_DESKTOP (zero footprint elsewhere), sitting beside
read_terminal/close_terminal in _HERMES_CORE_TOOLS.
Deliberately not adding run_slash: letting the agent fire slash commands
mid-turn (/model, /new, /clear) fights prompt-cache + conversation
invariants.
Plain Enter (and the primary send button) while a turn is running now redirects
the live turn with the typed correction instead of queueing it — matching
Cursor's stop-and-correct. Removes the now-redundant steering-wheel button
(redirect is the default gesture) and teaches the primary button the `steer`
action. Attachments still queue; slash commands still run inline.
Update the CLI and messaging guides so the default `interrupt` mode reflects
the new behavior: a follow-up redirects the active turn (preserving displayed
reasoning and completed work, letting running tools finish at a safe boundary)
rather than hard-stopping it, with `/stop` still the explicit hard stop.
The default `busy_input_mode: interrupt` now redirects the live turn instead
of hard-stopping it and re-queuing a fresh turn, wired consistently across
every first-party surface via the shared core primitive.
- CLI, gateway (busy + PRIORITY paths), TUI (`_handle_busy_submit`), desktop
(`session.redirect` RPC), and ACP call `redirect()` when the agent advertises
`_supports_active_turn_redirect`, and fall back to the proven interrupt +
next-turn queue for older runtimes.
- Redirect is gated to plain text with no attachments: captioned or
attachment-bearing events (including adapters that classify unknown media as
`TEXT`) stay queued so media is never dropped.
- ACP `cancel()` records the interrupted prompt, sets its cancel event, and
hard-stops the agent while holding `runtime_lock`, closing the
cancel-then-correct ordering gap; connection I/O happens after the lock is
released.
- Desktop appends the correction as a real user transcript message so the live
view matches the durable history after reload.
- `/busy` help, onboarding hints, and the new `session.redirect` RPC describe
the redirect behavior; `/stop` remains the hard stop.
Add a desktop-gated open_preview tool so 'open cnn.com in the preview
pane' works. The tool (check_fn on HERMES_DESKTOP, zero footprint
elsewhere) emits a preview.open event through a gateway-injected emitter,
mirroring the close_terminal -> terminal.close bridge. The desktop
handles it in usePreviewRouting, normalizing the target and opening the
pane for the active session only -- a background turn never hijacks it.
Bare domains and localhost are coaxed into fetchable URLs (www.cnn.com ->
https://, localhost:3000 -> http://); file paths and schemes pass through
to the renderer's normalizer.
The Codex app-server runtime bypasses the main conversation loop and drives
its own subprocess turn, so it needs first-class hooks rather than the
OpenAI-loop interrupt path.
- `AIAgent.interrupt()` now forwards a hard stop to
`CodexAppServerSession.request_interrupt()`, and `redirect()` uses Codex's
native `turn/steer` protocol instead of cancelling the subprocess.
- `run_turn()` no longer clears an interrupt that arrived during
`ensure_started()`: a stop landing mid-startup is honored before `turn/start`,
and the interrupt event is cleared on every exit path.
- `run_codex_app_server_turn()` mirrors the loop finalizer's interrupt handoff
(surface `interrupted` / `interrupt_message`, then `clear_interrupt()`) on
both the normal and exception early-return paths, so a hard stop can't leave
`_interrupt_requested` stale for the next turn.
A follow-up sent while the model is still generating previously ended the
turn: Hermes kept only the visible partial text (reasoning was display-only),
cleared the loop, and replayed the message as a fresh next turn. If the
correction referred to something that only appeared in the thinking stream,
the model no longer had that context.
Add `AIAgent.redirect(text)`: a corrective interrupt distinct from a hard
stop. It cancels only the in-flight model request (not tool workers or child
agents), stashes the correction under a lock shared with `interrupt()` so a
concurrent `/stop` always wins, and lets the loop rebuild the same logical
iteration. `_apply_active_turn_redirect()` checkpoints the reasoning that was
actually shown to the user plus any visible partial text as an ordinary
assistant message, then appends the correction as a real user turn — never
replaying incomplete signed/encrypted provider reasoning, and keeping strict
role alternation and prompt-cache stability intact. During tool execution it
degrades to `steer()` so a running tool finishes at a safe boundary.
`_fire_reasoning_delta` now only records reasoning that a display callback
actually consumed, so `show_reasoning: false` never leaks hidden provider
thinking into the persisted transcript.
Since #65919 the live view seals each chunk of mid-turn assistant
commentary (message.interim) as its own finalized bubble. Every bubble
with visible text renders the hover action footer, so a tool-heavy turn
grew a copy/refresh bar under almost every paragraph — and the live
render didn't match rehydration, which merges the turn into one bubble.
Mark sealed interim bubbles with ChatMessage.interim, carry the flag
into the runtime message metadata (custom.interim), and skip the
AssistantFooter for them. The turn's final reply keeps the footer; a
previewed final that settles onto an interim bubble clears the mark so
the settled reply regains its actions. interim joins COMPARED_FIELDS /
chatMessagesEquivalent so flipping it repaints.
Also fix an id-collision flake this surfaced: stream/interim bubble ids
were Date.now()-only, so an interim seal and the next segment's first
delta in the same millisecond reused the id and the new segment appended
into the sealed bubble. Ids now include a monotonic sequence.
Follow-up for the salvaged #55800 idle-compaction commit:
- turn_context.py: treat a skipped _compress_context (per-session
compression lock held by another path, failure cooldown, anti-thrash
breaker, codex-native routing) as a strict no-op — only re-baseline
conversation_history and re-anchor current_turn_user_idx after a REAL
compaction. Also re-anchor the user-message index after idle compaction
(the PR predates the reanchor helper).
- hermes_cli/config.py: add idle_compact_after_seconds: 0 to
DEFAULT_CONFIG's compression block (the PR only had
cli-config.yaml.example).
- gateway/run.py: add the idle-compaction status wording to
_TELEGRAM_NOISY_STATUS_RE so the new 💤 message stays out of
human-facing chat surfaces (routine compaction is silent by design);
pin it in tests/gateway/test_telegram_noise_filter.py.
- docs: idle_compact_after_seconds in user-guide/configuration.md and
developer-guide/context-compression-and-caching.md parameter table.
- tests/agent/test_idle_compaction_lock_and_guards.py: end-to-end
coverage with a real AIAgent + SessionDB proving the idle path honors
the per-session compression lock (added after the PR), the persisted
failure cooldown, and the anti-thrash breaker, and that the lock is
released after an idle-triggered compaction.
Salvaged from #55800 by @iso2kx. Implements #27579.
Long-lived sessions (e.g. a Telegram thread resumed over hours/days)
accumulate a large context that the existing size-based threshold only
trims once it crosses `threshold × context_window`. Until then every
turn re-reads the full history, which on large-context models can mean
hundreds of K of cache-read tokens per call even across long idle gaps.
Add a time-based trigger that complements (does not replace) the size
threshold: when a session resumes after `compression.idle_compact_after_seconds`
of inactivity, compact the accumulated history up front, before the first
reply. Disabled by default (0), so existing behaviour is unchanged.
The trigger reuses `_last_activity_ts` (the last time the turn loop did
work) to measure the idle gap at turn start, gates the token estimate
behind a cheap gap pre-check, and skips compaction when the context is
already at/below the post-compression target (threshold × target_ratio)
so a short idle thread never pays for a summarization that saves nothing.
It also defers to an active compression-failure cooldown.
The decision is factored into a pure predicate, `_should_idle_compact`,
which is unit-tested without a live agent.
Composing #57835's multi-fossil summary scan with #47274's merged-handoff
unwrap: when the restart-decay path pulls a merged handoff into the
compression window, its genuine prior-tail user content must enter the
summarizer input (folded into the fresh summary) rather than being
dropped with the summary row. Standalone handoffs still drop. The
continuity test now pins the composed contract: recovered verbatim OR
via summarizer input, never silently deleted, never duplicated.
- Three '_previous_summary == "fresh summary"' exact pins and one
transcript-wide fossil-absence pin predated main's deterministic
task-snapshot grounding (761a0b124e), which prepends a
'## Historical Task Snapshot' section to stored summaries and may
quote a folded head turn inside the handoff. Re-pin the contracts
(fresh body present, fossil absent from non-summary messages)
instead of exact strings.
- Add test_restart_simulation_fresh_compressor_does_not_reprotect_head:
a fresh ContextCompressor over a transcript containing a persisted
handoff summary computes the same decayed protected-head boundary
(compress_start base) as a live already-compacted process, and the
first post-restart compaction does not preserve pre-restart head
fossils (#57814).
protect_first_n decay state (compression_count / _previous_summary) is
in-memory only, so a gateway restart re-protected the persisted handoff
summary and head fossils, growing the head unboundedly across
restart+compaction cycles (#57814).
_effective_protect_first_n now probes a bounded resumed-head window for
a persisted handoff summary (by metadata or content prefix) and decays
protection when one is found, before compress_start is computed. The
first post-restart compaction self-heals: stacked summary fossils are
folded into the next summary prompt instead of preserved verbatim,
rehydration is rolled back on abort, deterministic fallback carries a
bounded redacted previous-summary snapshot, and forced user-leading
merged summaries keep the live tail request after the summary end
marker so they stay rehydratable.
Squashed reapply of the 12-commit series from PR #57835 onto current
main (branch was 1210 commits behind; single add/add conflict with the
task-snapshot grounding helpers resolved by keeping both).
Fixes#57814.
Cleanup pass on the per-request provider-options merge — behavior
unchanged:
- collapse the duplicated entry-validation shared by merge() and its
callers into a single guard inside merge()
- read the configured memory provider in readable steps instead of a
nested ternary
- build the merged mapping as one {**base, **overlay} expression
- space out logical blocks
Addresses review on #69077. The first pass added a second, heavier
round-trip (`GET /api/memory` -> `_discover_memory_provider_statuses()`,
which imports every provider module and probes install state) just to
fill the desktop dropdown, and left `schema.options` for memory.provider
dead — three sources of truth for one list.
Root cause is narrower: the desktop schema *already* carried a
discovery-driven `memory.provider` option list (`_SCHEMA_OVERRIDES` ->
`_memory_provider_options()`), but `enumOptionsFor` returned the static
`ENUM_OPTIONS['memory.provider']`, which shadowed `schema.options` in
config-field.tsx. The only real gap was liveness: `_SCHEMA_OVERRIDES` is
frozen at import time, so a provider installed mid-session never showed.
Fix at the layer the rest of this stack already uses:
- Backend: generalize `_schema_with_voice_provider_options` ->
`_schema_with_dynamic_provider_options`, which now also recomputes
`memory.provider` options per request (cheap plugin-dir scan via
`_memory_provider_options`, plus current-value preservation). Fixes the
same staleness for CLI + dashboard, not just desktop.
- Frontend: drop the `memory.provider` entry from `ENUM_OPTIONS` so
`enumOptionsFor` returns undefined and config-field consumes the
discovery-driven `schema.options` directly. No new frontend round-trips.
- Remove the now-unnecessary `getMemoryStatus()` fetch/state/wiring in
config-settings.tsx (reverted to main).
- Fix the stale `helpers.ts` comment ("schema omits memory.provider").
Tests: backend tests for the per-request merge (recomputes discovered
providers; preserves a configured-but-undiscovered value); frontend test
asserts enumOptionsFor no longer shadows the schema for memory.provider.
Co-authored-by: brooklyn! <770929+OutThisLife@users.noreply.github.com>
The desktop Settings memory-provider dropdown read a hardcoded
`ENUM_OPTIONS['memory.provider'] = ['', 'honcho', 'hindsight']` list,
so user-installed and pip-installed providers never appeared even though
the backend already discovers them (`GET /api/memory` ->
`_discover_memory_provider_statuses()`) and the CLI (`hermes memory
setup`) lists them. This was the one surface left where the memory
config stack was not schema/discovery-driven.
Fetch `getMemoryStatus()` on the settings page (mirroring the existing
`elevenLabsVoiceOptions` pattern) and pass the discovered provider names
to `enumOptionsFor` as `dynamicOptions` for the `memory.provider` key.
The static `ENUM_OPTIONS` entry is demoted to a pre-load fallback; the
current-value passthrough still keeps a selected-but-undiscovered
provider visible.
Completes the desktop half of the schema-driven memory-provider config
surface (the CLI + backend + generic panel already landed via #51020 /
#67206), superseding the stale #48675 which built the same feature
against the pre-refactor layout.
Co-authored-by: brooklyn! <770929+OutThisLife@users.noreply.github.com>
2026-07-21 23:13:50 -04:00
1370 changed files with 115979 additions and 13865 deletions
echo "- [playwright-report]($REPORT_URL) — interactive HTML report" >> "$GITHUB_STEP_SUMMARY"
echo "- [playwright-report]($REPORT_URL) — interactive HTML report"
fi
if [ -n "$DIFFS_URL" ]; then
echo "- [visual-diffs]($DIFFS_URL) — just the diffed screenshots (small, fast to review)" >> "$GITHUB_STEP_SUMMARY"
echo "- [visual-diffs]($DIFFS_URL) — just the diffed screenshots (small, fast to review)"
fi
echo "" >> "$GITHUB_STEP_SUMMARY"
echo "**To update baselines:** merge to main (baselines auto-update on main runs) or run \`npx playwright test --update-snapshots\` locally." >> "$GITHUB_STEP_SUMMARY"
echo ""
echo "**To update baselines:** merge to main (baselines auto-update on main runs) or run \`npx playwright test --update-snapshots\` locally."
# Also parse the JSON report for pass/fail counts
if [ -f playwright-report/results.json ]; then
echo "" >> "$GITHUB_STEP_SUMMARY"
echo "### Test Results" >> "$GITHUB_STEP_SUMMARY"
echo "" >> "$GITHUB_STEP_SUMMARY"
echo ""
echo "### Test Results"
echo ""
node -e "
const r = require('./playwright-report/results.json');
console.error("::error::Workspace discovery produced an empty package list — refusing to emit a zero-length matrix (would skip all JS/TS checks silently).");
"skill with skill_manage so you can reuse it next time.\n"
"When using a skill and finding it outdated, incomplete, or wrong, "
"patch it immediately with skill_manage(action='patch') — don't wait to be asked. "
"Skills that aren't maintained become liabilities."
"Skills that aren't maintained become liabilities.\n"
"\n"
"## Skill Safety Rule\n"
"1. **UNAVAILABLE** — If a skill placeholder contains `[SKILL_PRUNED]`, the skill content was lost in compression and is inaccessible.\n"
"2. **RELOAD** — Before performing any action that depends on a skill, re-check its content with `skill_view(name='...')` if it shows `[SKILL_PRUNED]`.\n"
"3. **WAIT** — If a skill is loading or was just pruned, wait for the reload confirmation before proceeding.\n"
"4. **DEDUP** — After reloading a pruned skill, **ignore any remaining `[SKILL_PRUNED]` markers for that same skill** — they are historical artifacts from previous compactions and do not need further action."
)
KANBAN_GUIDANCE=(
@@ -884,7 +890,14 @@ PLATFORM_HINTS = {
"You're responding through an API server. The rendering layer is unknown — "
"assume plain text. No markdown formatting (no asterisks, bullets, headers, "
"code fences). Treat this like a conversation, not a document. Keep responses "
"brief and natural."
"brief and natural."
"File/media delivery: images referenced as MEDIA:/absolute/path tags "
"(.png/.jpg/.jpeg/.gif/.webp/.bmp, up to 5MB) are inlined as base64 data "
"URLs in responses on the chat, completions, and responses endpoints. "
"Non-image files are NOT intercepted anywhere, and the runs endpoint "
"intercepts nothing — a MEDIA: tag there renders as literal text exposing "
"a raw host filesystem path. For those cases, state the plain file path "
"in your response text instead of a MEDIA: tag."
),
"webui":(
"You are in the Hermes WebUI, a browser-based chat interface. "
@@ -30,7 +30,7 @@ Already have the Hermes CLI? Just run:
hermes desktop
```
It builds and launches the GUI against your existing install — same config, keys, sessions, and skills. On first launch Hermes walks you through picking a provider and model; nothing else to configure.
It builds and launches the GUI against your existing install — same config, keys, sessions, and skills. If Desktop cannot find a usable runtime or saved remote connection, first launch lets you connect to an existing Hermes gateway or install Hermes locally. Local onboarding then walks you through choosing a provider and model.
### Prebuilt installers
@@ -134,6 +134,19 @@ Desktop supports a managed local backend, explicit remote gateways, and Hermes
Cloud connections. Remote and cloud modes use the same remote-capability path;
authentication and discovery differ, not the renderer feature model.
When no usable local runtime or saved remote connection exists, the first-run
screen offers **Connect to existing Hermes** before starting the local installer.
Desktop probes the gateway to discover token or OAuth authentication, requires a
successful HTTP and WebSocket connection test, and saves the connection using
the same encrypted Desktop configuration used by Settings. A saved remote
connection bypasses this choice on later launches. The regular Desktop build
still includes the local-install option; this is a remote operating mode, not a
separate client-only application.
In remote mode the gateway host is the execution boundary: agent tools,
terminal commands, and file operations run against the remote Hermes host, not
the computer displaying the Desktop UI.
Projects are the workspace abstraction. A project may own multiple folders,
repositories, worktrees, and sessions; a bare new chat remains detached unless
the user enters a project or configures a default project directory. Use the
consttimer=setTimeout(()=>reject(newError(`Timed out after ${DEFAULT_TIMEOUT_MS/1000}s waiting for ${operation}:\n${this.stderr.join('\n')}`)),DEFAULT_TIMEOUT_MS)
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
bi
