Commit Graph
64 Commits
Author SHA1 Message Date
andybpan 5f06e1b0a3 fix(core): treat files package managers run by default as tool-trusted (#8129)
npm, pnpm, yarn and bun run `node-gyp rebuild` on install when a binding.gyp
is present and package.json has no install script, and a .gyp file can include
any .gyp or .gypi. npm, pnpm and yarn run `node server.js` for start when
package.json has no start script. A project from `hyperframes init` has
neither script, so both defaults apply.

toolTrustedFileKind now classifies .gyp/.gypi files and server.js, in every
spelling projectPathSegments folds, so Studio's file routes and the registry
installer refuse to write them. Rename and duplicate refusals now report
why: protected_path when the destination is inside the project, as POST and
PUT already do.
2026-11-18 01:32:34 +00:00
terencecho 414adffb82 fix(skills): select standalone CLI releases through the launcher (#8102)
* fix(skills): select standalone CLI releases through the launcher

* fix(cli): clarify release commands and share launcher validation
2026-11-17 21:41:22 +00:00
Michael Wang 2a28e43685 fix(studio-server): no write route touches files other tools run or trust (#7964)
* fix(studio-server): no write route touches files other tools run or trust

A composition previewed in Studio runs same-origin with the API, and the write rule
refused only private paths, node_modules and hidden names. Its script could create,
overwrite, rename or delete package.json, lock files, agent instruction files and
scripts, which a package manager, build tool or coding agent then runs with the
developer's rights.

Core now names these files in one predicate, toolTrustedFileKind, read in every
spelling a case-insensitive file system or Windows opens. Studio's write rule and the
registry installer's target rule both use it, so every write route (PUT, POST, PATCH,
DELETE, upload, duplicate-file, file-mutations and gsap-mutations) refuses such a
target. Deleting or moving a folder that holds one is refused too. Ordinary
composition, stylesheet, script-asset and media writes still succeed.

* fix(core): fold look-alike letters before matching files other tools trust

A case-insensitive macOS volume opens "AGENTſ.md" (long s) as AGENTS.md
and "Makefile" (fi ligature) as Makefile, but lower-casing alone does not
map either letter. Project path segments now fold names the way such a
volume does: NFKC, upper-then-lower case, and invisible characters
dropped. Every rule built on the segments gets the fold.

Also:
- name JSON, YAML and TOML tool config, compose files, turbo.json,
  nx.json, pyproject.toml and gulp or grunt files
- treat a ~N name as a short name only when it has the 8.3 shape and an
  extension a trusted file's could shorten to, so take~2.mp4 stays
  writable
- refuse to delete or move a folder holding node_modules or .git, walking
  it without the listing's ignore rules
- report an upload refused by the write rule in `skipped`
2026-11-16 21:30:32 +00:00
Michael Wang 8959deb00f fix: snapshot skips private files, no text-art demuxers, escaped warning text (#7943)
* fix(engine): video warnings show control characters in the src as visible escapes

The missing-source and opaque-alpha warnings pass the authored src, and the path
built from it, through terminalText before writing them to stderr.

* fix(cli): snapshot never probes or decodes a private project file for a page video

projectLocalFile applies the same private-file rule as the render readers, judged as written
and with links resolved, so held-video probes and frame injection skip those files. Studio's
own prepared-assets and requests folders still resolve.

* fix(parsers,engine): media inputs never select a text-art demuxer

The media format allowlist leaves out tty, bin, xbin, idf and adf, so ffmpeg and ffprobe
cannot draw a text file as a video frame, whether or not ffprobe is installed. ffprobe's
refusal of one of these formats still reports the source as a text file.

A link inside the project whose target is outside it is still read by snapshot: links a
person puts in their own project are accepted, imported projects cannot carry links, and
render reads them the same way.

* fix(cli): registry warnings show control characters as visible escapes

A project's hyperframes.json can name the registry, so item names and deprecation text
from it pass through terminalText before the studio install, remote template and
item-loading warnings write them to stderr.

* fix(cli): failed render and frame read messages escape control characters

The render failure box escapes each line of the message, keeping its line breaks, and
snapshot's frame read errors escape the path and the ffmpeg or ffprobe text they quote.

* test(cli): probe a missing path with control characters instead of creating it

Windows cannot create a file with those characters in its name; the failed probe's
message still carries the path.
2026-11-16 18:13:39 +00:00
Michael Wang 6a7b4df7c8 fix(cli): add lists files from a received config's registry as received (#7820)
* fix(cli): add lists files from a received config's registry as received

When a project's hyperframes.json is listed in .hyperframes/received-files.json
and names a registry other than the public one, add now lists every file it is
about to write in that record before writing any of them. This covers every
caller of addToProject: add by name, by tag, several names, and Studio.

The decision reads the config the install uses, so a byte-order mark or another
spelling of the file on disk is handled the same way. A person's own config, the
public registry and a record that cannot be read are left as they are.

* fix(cli): list received files as they are written, under a shared lock

add now lists each file in the received-files record just before it writes
it, using the write-time decision, so a file kept because the person edited it
during the download is never listed. When an install fails, it takes back the
entries it listed but never wrote.

Changes to the record take .hyperframes/received-files.json.lock, created
exclusively. A writer retries for about 2 s, takes over a lock older than
about 10 s, and always releases it. A writer that cannot take the lock in time
fails with a clear error instead of writing past it.

* fix(cli): an unreadable config counts as received; a failed install keeps what is on disk

add now counts an install as received when hyperframes.json exists but cannot be
read or parsed, since that config is not proven to be the person's own. When an
install fails, it takes back only the entries it listed whose files are not on
disk; files that landed stay listed.

* fix(cli): the received-files lock resolves the project folder as the installer does

The record's lock and write resolved the project folder with fs.realpathSync,
while the containment check resolves targets with the native realpath. The two
can spell the same folder differently: a Windows short name or a folder typed
in another letter case. The target then looked outside the project and the
change was refused. The record now resolves the folder with the installer's
realProjectRoot, so the root and target compare in one spelling.

The failed-install tests now fail the write by stubbing it, which works on every
platform, instead of a read-only folder, which Windows ignores.
2026-11-15 09:10:08 +00:00
wonjangandClaude Fable 5.2 8dbbc1d371 fix(registry): caption keywords, emoji and aftershock work for words outside a-z (#7804)
caption-emoji-pop and caption-neon-accent look a word up in KEYWORDS and
WORD_EMOJI after stripping everything outside a-z, and caption-glitch-rgb
counts a word's \w characters to decide on the aftershock. A Korean, Cyrillic
or accented word comes out empty or cut, so a keyword written in that script
never matches and a long word never pulses.

The three components now strip by Unicode letter (\p{L}) instead, which is
the same set for a lowercased English transcript.

Co-authored-by: Claude Fable 5.2 <noreply@anthropic.com>
2026-11-15 05:44:00 +00:00
Miguel Ángel e21e64f6ab fix(registry): caption-highlight clears a short word's highlight when the word ends (#7735)
* fix(registry): caption-highlight clears a short word's highlight when the word ends

On a word shorter than 0.05 s the 0.15 s fade-in outlived the fade-out, so
a frame inside that tail lit the highlight again until the phrase ended.

* fix(registry): leave the generated caption-highlight catalog files to the generator
2026-11-14 14:39:00 +00:00
James RussoandClaude Opus 5.5 9bab57ab48 fix(cli): refuse more coding-agent instruction files in registry installs (#7549)
Registry items may not write QWEN.md, CRUSH.md, WARP.md or CONVENTIONS.md at
any depth, in any case, alongside the AGENTS.md, CLAUDE.md and GEMINI.md
names already refused. The catalog's nested SKILL.md and README.md notes are
unaffected.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-11 08:35:03 +00:00
James RussoandClaude Opus 5.5 b96c82b785 fix(cli,studio-server): limit registry installs and Studio file writes to project content folders (#7523)
* fix(cli,studio-server): limit registry installs and Studio file writes to project content folders

Registry blocks and components now install only into the project's
compositions, components and assets folders (plus models/ and lib/, which
the catalog uses, and notes at the root). node_modules, agent instruction
files and programs or scripts are refused at any depth, matched
case-insensitively with either separator, and checked again on the path
the project's own links lead to.

Studio's file routes never read or write git metadata or env files, and
every route that writes, renames, deletes, copies or uploads a
request-named path refuses the project folder itself, node_modules and
hidden files or folders other than .hyperframes/ and .media/, on both the
spelled and the link-resolved path.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* chore(cli,studio-server): trim comments added by the write allowlist

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli,studio-server): trim path segments in linear time

The trailing dot/space and :stream trims used anchored regexes that
backtrack quadratically on long runs of spaces, dots or colons in a
request path. Trim with a plain scan instead.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): refuse example dependencies in add and allow only TEMPLATE.md at the root

- addToProject now refuses a block or component that depends on an example,
  not only an example asked for by name. The installer lets an example write
  at the project root, so a dependency must not carry that through.
- The only file a block may write at the project root is TEMPLATE.md, the
  catalog's usage note. Any other .md or .txt name could be one a tool loads.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-11 07:53:52 +00:00
Michael WangandClaude Opus 5.5 84c7d02a19 fix(cli): keep project folders from choosing the programs and config the CLI runs (#7511)
* fix(cli): resolve helper programs from PATH, never the working directory

On Windows a bare command name is looked up in the working directory before
PATH. The CLI now sets NoDefaultCurrentDirectoryInExePath at startup, which
Node's spawn and cmd.exe both honor and child processes inherit, so tools
such as powershell.exe, taskkill and npx.cmd always come from PATH.

The shared ffmpeg/ffprobe resolver no longer scans the working directory on
Windows and skips relative PATH entries.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): keep Python helpers from importing files in the working directory

`python -c` and `python -m` put the working directory first on sys.path, so
a file in the project named like a probed module ran instead of the
installed package. Inline helper code now drops that entry before any
import, and `-m` helpers (pip) run through runpy after the same step. Paths
and arguments are unchanged.

On Windows, findPython also skips a `where` match in the working directory.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): run the skills installer and nested CLI calls independent of the project

The global skills install ran `npx skills@<pin>` in the current folder, so
npm could take the package or registry from that project's node_modules or
.npmrc. It now runs from the home folder; the install is global either way.

When the CLI starts the media-use engine it hands over its own entry file.
The local voice fallback and the audio helpers then run that same CLI with
the same Node instead of resolving `npx hyperframes` from the project.
Without a handed-over entry they behave as before.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(media-use): load only provider keys from a project .env

The audio scripts copied every key of a project .env into their environment
except two. They now take the same kind of short allowlist the CLI uses:
HeyGen, ElevenLabs and Gemini API keys plus a few harmless settings. Program
overrides, runtime options, credential files and the credential store path
come only from the person's own environment.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): build whisper.cpp only from a pinned, verified release

The source-build fallback cloned the upstream default branch. It now clones
the v1.9.5 release tag and refuses to build unless the checkout is the
reviewed commit for that tag.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): keep registry caches apart and refuse config install targets

Registry cache file names now include a digest of the exact registry URL,
and each entry records the URL it came from, so one registry's cached
manifest is never read for another. Entries from older versions are
refetched once.

Registry installs refuse targets inside .git or any hidden directory, hidden
configuration files, and package.json, hyperframes.json and
hyperframes.lock.json. An example installed by `init` may still bring its
own package.json and .gitignore.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(capture): keep page text to one line in agent files and validate Vertex keys

Capture writes AGENTS.md, CLAUDE.md and .cursorrules from the page's title,
font names and colours. Page text is now flattened to one short line and
colours must be hex values. Existing agent files that capture did not write
are kept, with a warning.

HYPERFRAMES_VERTEX_SERVICE_ACCOUNT must now be a plain service-account key
with a Google token endpoint; other credential types skip captioning with a
warning.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* test(cli): find the digest-named registry cache and trim a new file's comment

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): match registry targets as Windows names them; skills install without HOME

Windows drops trailing dots and spaces from a path segment and reads name:stream as
name, so package.json. or .git./hooks reached the files the target rules refuse.
The global skills install now falls back to the temp folder when HOME does not exist.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): turn off the Windows working-directory lookup in every program entry

The warm-check server and client are their own tsup entries and copied the
environment before cli.js set NoDefaultCurrentDirectoryInExePath, so the server
spawned a bare powershell.exe that Windows could take from the project folder.
Every entry that hides child consoles now also loads windowsExeSearch first.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): refuse Windows short-name registry targets

A segment like CLAUDE~1 or PACKAG~1.JSO opens the long name on NTFS, so it
reached the files the target rules refuse. The trailing-dot and stream-name
normalization is removed: publication already rejects those segments.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-10 20:01:19 +00:00
Miguel Ángel b59fa20833 fix(cli): staging folders beside a project path no longer fail past 260 characters on Windows (#7338)
* fix(cli): staging folders beside a project path no longer fail past 260 characters on Windows

mkdtemp does not get the long-path prefix that mkdir does, so every staging folder
made beside a project or output path failed once its full path passed 260 characters.
They now come from one helper, createStagingDir in @hyperframes/core/atomic-file, and a
scan test fails on a new mkdtemp beside a path.

* fix(cli): give mkdtemp the long-path form instead of replacing it

CodeQL treats mkdtemp as the safe way to make a folder under the system temp dir and
flagged every file written into a folder from the mkdir-based helper. createStagingDir
now calls mkdtemp with the long-path form of the path on Windows, so both hold.

* fix(cli): leave non-Windows paths alone when making a staging folder

The long-path form is now chosen by the shape of the path, not process.platform,
so tests that pretend to be Windows keep working and the comment share does not rise.

* fix(core): a staging folder keeps a Windows path that already names its namespace

* fix(core): a staging folder keeps a Windows device path as the caller gave it

* fix(core): the staging folder helper's comment fits the line limit
2026-11-09 17:16:08 +00:00
Miguel Ángel ecf0e14387 fix(cli): previews and every per-user cache follow HYPERFRAMES_CACHE_DIR (#7224)
* fix(cli): a preview's session follows HYPERFRAMES_CACHE_DIR when it is set

* docs(cli): background previews follow HYPERFRAMES_CACHE_DIR

* test(cli): unstub the preview test's env in a block body

* test(cli): clear HYPERFRAMES_CACHE_DIR in the CLI's test home

* fix(cli): registry cache, search model, catalog and catch-up follow HYPERFRAMES_CACHE_DIR

* style(cli): format the media search cache test

* test(cli): compare the model path against an absolute cache folder on Windows too

* fix(cli): both catalog readers treat an empty HYPERFRAMES_CATALOG_ARTIFACT_DIR as unset
2026-11-05 10:56:35 +00:00
WaterrrForeverandClaude Opus 5.5 fb738015a0 fix(cli): bug-bash fixes for pull, offline, preview, check, cloud aspect, sfx search, and media-use avatars (#7170)
* fix(cli): an interrupted templates pull leaves no partial folder

Unpacking now runs tar asynchronously and removes its .part folder on
SIGINT/SIGTERM/SIGHUP before re-raising the signal. Part folders carry the
owning pid, and the next pull clears any whose process is gone.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): auth status and templates pull say when HeyGen cannot be reached

Offline, both commands waited on unbounded requests and then printed only
undici's 'fetch failed'. A new networkFailure helper names the host, the
reason (DNS lookup, no route, refused, no answer in time, dropped) and the
next step. The account and token-refresh requests now time out after 15 s,
the template link after 15 s, and auth status reports an offline refresh at
once instead of retrying with the stale token. describeCauseChain moves into
the helper so the registry and auth share it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): preview leaves the project's HTML files as they are

Opening a project in preview stamped data-hf-id into every composition on
disk. The preview route and every source mutation already mint the same
content-keyed ids in memory, so the startup pass is dropped: files gain ids
only when an edit to them is saved. The .hyperframes folder is still made
before the watcher starts, so the preview cache creating it is not reported
to Studio as a manifest change.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): check runs the browser audits past a lint error

A lint error made check skip runtime, layout, motion and contrast, so one
known lint error hid every browser finding. The browser pass now runs
whenever there is a composition to load, and the lint error still fails the
check. Only a root with no data-composition-id skips it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): no desktop-app line for CI or a script reading render output

The 'Keep editing by chatting with Framey' line printed on every render and
preview, including CI and piped runs where nobody can open the app. It now
prints for a person at a terminal, and for a coding agent, which the
hyperframes skill tells to pass the line on; CI and other non-TTY runs get
nothing.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): cloud render takes a composition at any aspect ratio

A 1920x960 composition got 'server will default aspect_ratio to 16:9'
although the render kept 1920x960: at 1080p the cloud renders at the
composition's own size and aspect_ratio only picks a 4K preset. The CLI now
names the detected ratio (2:1) and says the size is kept, takes
--aspect-ratio as any W:H checked against the composition, sends only the
16:9, 9:16 and 1:1 the API accepts, and refuses a 4K render of any other
ratio before upload, where no preset fits.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(media-use): a weak sound-effect match says so instead of passing silently

resolve --type sfx returned a HeyGen sound at any score down to 0.4, and a
project file that shared one word of its name with the intent, with nothing
to tell them from a real match. Such a result now carries a match object in
the JSON ({score, min_score, low_confidence} or the matched words and the
intent the file was resolved for) and a low-confidence line on stderr. A
HeyGen search with no hit says so with the min score it used, and a sound
effect nothing matches fails with 'no sound effect matched'.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* feat(cli): media-use avatars lists the film account's avatar looks

Templates had to reach for 'heygen avatar list', a separate sign-in that can
be another account. 'media-use avatars' lists the private looks on the
HyperFrames account making the film (GET /v3/avatars/looks?ownership=private,
following next_token up to --limit) as {avatars: [{avatar_id, name, ...}]},
the id that 'generate --type avatar --avatar' takes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* chore(skills): leave style-sampling's avatar line to #7164, which owns that skill

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): templates pull removes only the part folders it marked as its own

The abandoned-part cleanup deleted any sibling named '<slug>.part-*' whose
next token was not a live pid, so a person's own 'sample.part-notes/',
'sample.part-reference.txt' or a symlink went with it. A pull now writes an
ownership marker (owner tag + pid) into its part folder, and the cleanup
removes only a real directory with the generated '<slug>.part-<pid>-XXXXXX'
name whose marker names a dead pid. Files, symlinks, unmarked folders and
older unmarked parts are left alone.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): media-use avatars says it is offline when the login refresh cannot reach HeyGen

A network failure while refreshing the stored login was reported as 'no
HeyGen sign-in' with 'auth login' as the fix, though the login was still
there. It now keeps the network failure's message and next step.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(media-use): a reused weak sound keeps its low-confidence advisory

The first resolve of a weak HeyGen sound or a name-matched project file
flagged it, but an identical second resolve hit the project manifest and
returned it with no match object. The advisory is now rebuilt from the
record's provenance (the stored score, or the adopted file's name) on reuse.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* refactor(cli): keep the comment share of files this branch touched from rising

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): no check-then-use on pull markers, and the download is never written to disk

CodeQL flagged the downloaded bundle written to a file before tar read it,
and a stat-then-read in the preview test. The verified archive now streams
into tar's stdin, so no network bytes reach a file the CLI writes. The
abandoned-part scan takes the entry type from readdir instead of an lstat,
reads the ownership marker through one O_NOFOLLOW descriptor it fstats, and
the preview test reads each file once instead of stat-checking it first.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* chore(ci): raise the CLI's unpacked size budget to 5% over its measured size

main's @hyperframes/cli was already about 0.99% over its 23722874-byte
unpacked budget, so this branch's additions (networkFailure, the avatars
command, the larger skill references) tip it past the 1% tolerance. This is
the same value #7169 sets, 25159107 bytes (5% headroom over the measured
size), so the two PRs merge cleanly.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* refactor(cli): split the pull-marker check and share the token-endpoint POST

Fallow flagged pullOwner (CRAP 49.5, cyclomatic 13) and a 28-line clone
between refreshOnce and exchangeCodeForTokens. pullOwner is now the name
match, the O_NOFOLLOW marker read and the marker check as three small
functions, with the same behaviour. Both token requests go through one
postToTokenEndpoint, which keeps the 15 s timeout and the named network
failure.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(cli): a 4k cloud render checks the exact size against its preset before upload

The aspect detector matches a ratio within ±0.05, so at 4k a 1920x1055
composition was sent as 16:9 and 1000x980 as 1:1, and the producer then
refused both as an aspect mismatch. At 4k the matched ratio's preset
(landscape-4k, portrait-4k, square-4k, the ones the API maps it to) now goes
through the producer's own checkOutputResolutionCompatibility, and a
composition it does not fit is refused locally with the producer's reason
and the 1080p alternative. 1080p, where the render keeps the composition's
own size, is unchanged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-04 18:21:36 +00:00
Miguel Ángel 5a07ea2e5f fix(cli): catalog says the registry could not be read instead of listing nothing (#7015) 2026-11-02 14:47:04 +00:00
Miguel Ángel 47a1a92064 fix(cli): a missing registry item names the closest ones, and a blocked browser says how to allow it (#6799)
* fix(cli): a missing registry item names the closest ones, and a blocked browser says how to allow it

* fix(cli): let closestNames carry its rule in its name
2026-10-30 06:15:03 +00:00
Miguel Ángel 32c01c99a1 fix(registry): caption components keep the last phrase until its last word ends (#6792)
* fix(registry): caption components keep the last phrase until its last word ends

* test(registry): caption last group stays visible until its last word ends

* test(registry): narrow the timeline lookup for strict typecheck
2026-10-30 06:07:40 +00:00
Miguel Ángel e21fd123fd fix(cli): hyperframes add keeps going when Windows holds the file it replaces (#6576)
Publishing a registry file (every block file and hyperframes.lock.json) renamed
its staged copy over the destination with a plain rename, which Windows refuses
with EPERM while antivirus or a just-exited process still holds the file. It now
uses the shared rename that waits and retries on those errors.
2026-10-27 19:09:48 +00:00
Abdullah M. Nazif 9a517560a8 fix(cli): refresh item manifests when bypassing registry cache (#5256)
* fix(cli): refresh item manifests when bypassing registry cache

* chore: prepare conflict resolution against current main

* fix(cli): preserve registry refresh with bounded timeout fallback
2026-10-25 00:20:54 +00:00
Miguel Ángel d898ad89ee perf(cli): commands start sooner by compiling registry schemas on first use (#5935)
* perf(cli): commands start sooner by compiling registry schemas on first use

* fix(cli): registry validation keeps its file comment-free
2026-10-20 07:38:59 +00:00
Miguel Ángel 95a686c23d fix(cli): catalog loads registry items a few at a time so a slow link no longer times them out (#5829)
* fix(cli): catalog loads registry items a few at a time so a slow link no longer times them out

* chore(cli): keep the moved concurrency helper's doc to one line

* chore(cli): wrap the concurrency helper's doc under the line limit
2026-10-18 08:36:30 +00:00
Miguel Ángel 63a7ec692f feat(registry): eight more WebGPU shader backgrounds and a bundle with thirty new shaders (#5191)
* feat(registry): eight more WebGPU shader backgrounds and a bundle with thirty new shaders

* refactor(registry): simpler bundle copy step in the shader build

* test(registry): every shader block ships the same bundle and names a shader in it

* refactor(registry): one block loader for the shader block tests

* test(registry): name the stale block and match only shader names

* chore(registry): shorter licence line at the top of the shader bundle

* test(registry): every shader block installs the bundle's licences
2026-10-09 08:52:03 +00:00
Miguel Ángel b50dead1da feat(registry): gooey split block with metaball goo and directional motion blur (#5123)
* feat(registry): gooey split block with metaball goo and directional motion blur

A dot splits into a three-dot menu through a gooey neck, spins into a
swirl mark under shutter motion blur, breaks back into three dots and
merges into one. Timing, spacing and eases are measured off a reference
recording frame by frame.

The motion-blur snippet gains an opt-in `sharp: 0` option that hides the
crisp frame-time copy while an element moves, leaving only the shutter
average. The default is unchanged; the inlined copies carry the same text.

* feat(registry): grow gooey-split blades out of swelling dots

* refactor(registry): tidy gooey-split pose code and document sharp 0 limits

* fix(registry): hide a sharp 0 motion-blur target by marker, not its visibility

Hiding through the element's own visibility fought timelines that
write it (autoAlpha): a target faded in after attach stayed hidden at
rest. A marker attribute and one !important rule now do the hiding, the
default path writes nothing, and a declarative sharp other than 0 or 1
is refused like any other bad option.

* fix(registry): smear a motion-blur target that was hidden when blur attached

Copies froze the target's computed visibility at attach, so a target
entering with autoAlpha never smeared (and vanished entirely under
sharp 0). The group now carries the target's current visibility each
frame, read with the sharp 0 marker lifted, and a copy keeps a
descendant's visibility only where it differs from its parent's.

* test(registry): pin a copy keeping a descendant's own visibility

* fix(registry): keep a child's own hiding in the smear of a target hidden at attach

* refactor(registry): restore a lifted visibility in one call and lift any hidden form

* fix(registry): restore a lifted visibility even when copying styles throws

* fix(registry): let motion-blur copies hide with a clip that has ended

The smear group set an explicit visibility on every moving frame, and
copies used to freeze an explicit visible, so an ancestor clip hiding
after the last update could not hide them: copies from a 0-4 s
sub-composition stayed on screen at 16.4 s. The group now sets a
visibility only where the element's differs from its parent's.
Also regenerates skills-manifest.json for the motion-blur skill doc.
2026-10-06 21:24:38 +00:00
Miguel Ángel 590c192498 feat(catalog): add 3D objects, cursors, snow, starfield and paper transitions (#4349)
* feat(catalog): shelve 3D objects and cursors in their own catalog sections

Carries the generator grouping rules from #4310 and #4308.

* feat(registry): add a starfield ambient overlay

From #4313.

* feat(registry): add a falling snow ambient overlay

From #4331.

* feat(registry): add a torn-paper-rip transition

From #4321.

* feat(registry): add a paper-collage-reveal transition

From #4322.

* feat(registry): add a curtain-reveal transition

From #4323.

* feat(registry): add 3D objects: product turntable, glass hero, exploded stack, extruded logo

From #4310.

* feat(registry): add eight cursor animations and the dart cursor

From #4308.

* fix(catalog): find items by everyday editor words and accept a bare search word

From #4311.

* chore(registry): remove the simulated-cursor component

The Cursors section's new pointer animations replace it (from #4308).

* fix(registry): take cursor-zoom-follow's screenshot from runtime variables only

The declared default is empty, so page markup never becomes an image URL.

Clears the code-scanning alert raised on #4308.

* docs(cli): show a catalog search example that finds items

"smoke" matches nothing in the catalog; "crossfade" finds fade-through.

* chore(ci): allow the simulated-cursor removal in the deletion guard

* fix(ci): skip tracked files missing from disk in the player CDN pin check

Catalog regeneration removes a deleted item's pages before the publish PR untracks them,

so the check crashed on every catalog deletion PR.

* chore(cli): drop comments that narrate catalog search history

* fix(registry): drag-drop keeps the family arrow and shows the drop slot only while hovering

* fix(registry): agent cursor labels its steps in an accent tag, no glow or sparkles, in 4.8 s

* fix(registry): annotate cursor draws a tapered marker stroke and clears the circled value

* fix(registry): show real file thumbnails and end the context menu demo sooner

* fix(registry): give marquee select real photos, a Finder-blue marquee and a tighter loop

* fix(registry): centre the text-select toolbar on the selection and tighten its timing

* fix(registry): selection cursors default to the macOS system blue

* style(registry): format the agent and annotate cursor sources

* fix(registry): dart-cursor stops covering the choice it makes and shows the export finishing

* fix(registry): frame the product turntable fully and light it with two studio softboxes

* fix(registry): touch-indicator taps like a real phone and the post it taps gets liked

* fix(registry): press-ripple confirms with a check and small click rings at the cursor

* fix(registry): exploded stack separates at once and marks each layer in blue as it is labelled

* fix(registry): extruded logo settles facing the camera, set in Geist Bold with a neutral rim

* fix(registry): magnetic cursor docks its toolbar on the selection and ends after the press

* style(registry): format the exploded stack source

* fix(registry): glyph trail becomes a tutorial cursor path with numbered steps on a settings screen

* fix(registry): make glass-hero a clear lens that shows the headline through it

* fix(registry): snow reads as real snowfall with bokeh flakes over a winter title card

* fix(registry): oversized cursor is a 3.5x arrow that clicks a real editor toggle on

* fix(registry): starfield reads as a drifting night sky over a real title card

* style(registry): format the glass hero source

* fix(registry): make the paper tear show a white fibrous edge and lift the torn sheet away

* fix(registry): curtain reveal draws velvet folds that gather onto a title card

* fix(registry): make every zoom-follow click change the screen and read at zoom

* fix(registry): paper collage pieces stay opaque, tear with deckled grain and lift away on an arc

* refactor(registry): drop dead code and correct stale comments in reworked items

* fix(registry): agent cursor scrubs cleanly, docs match press-ripple, fictional demo names

* docs(registry): catalog notes say oversized-cursor exits and how it differs from press-ripple

* docs(registry): plain lowercase comments in oversized-cursor and touch-indicator

* docs(registry): oversized-cursor names CLICK_AT as the click cue
2026-09-28 22:03:59 +00:00
Miguel Ángel 26bfd38136 fix(studio-server): a save never recreates a project folder renamed away (#4602)
* fix(studio-server): a save never recreates a project folder renamed away

* style(cli): wrap imports after merging main

* fix(studio-server): proxy and upload answer that the project folder is gone instead of 502 or an empty 201

* test(studio-server): rename the project folder while an upload is being read

* fix(studio-server): a proxy or backup answers that a vanished project folder is gone

* test(studio-server): clean up the renamed folder after the proxy test

* refactor(studio-server): drop an unused re-export and share the folder-gone test setup

* fix(core): no folder outside a project is created once the project folder is gone

* test(studio-server): a refused render leaves no outside renders folder behind

* docs(core): wrap the mkdirWithinProject comment under the line limit

* test(studio-server): pass the project folder to the preview document store
2026-09-28 09:48:22 +00:00
Miguel Ángel 59e3850f6a fix(studio-server): a write through a link lands on the file the system resolves (#4589)
* fix(studio-server): a write through a link lands on the file the system resolves

* fix(cli): resolve catalog install paths as the system does, matching the install root

* fix(cli): list installed files relative to the real project folder
2026-09-27 12:30:17 +00:00
Miguel Ángel 2b1bb8b731 test(cli): CLI tests run with a home folder of their own, never the user's (#4587)
* test(cli): CLI tests run with a home folder of their own, never the user's

* style(cli): format the test home check

* test(cli): the skills test lane also runs with a home folder of its own

* style(cli): format the test home check

* ci: the test reachability map follows the skills lane's new command
2026-09-27 12:30:03 +00:00
Miguel Ángel 4984a268f7 fix(cli): Studio's catalog install records the item and honours paths, as add does (#4573)
* fix(cli): Studio's catalog install goes through add, so it records the item and honours paths

* fix(cli): Studio's catalog lists the project's registry and install paths survive symlinks

* refactor(cli): name the Studio catalog's registry option plainly

* feat(cli): registry_item_added says whether add or Studio installed the item

* fix(cli): dropping the same catalog block in Studio twice installs it again

* chore(cli): drop a comment that restates add's source option

* fix(cli): Studio's re-record finds lock entries by their resolved path

* test(cli): the symlinked-folder install test checks the resize happened

* fix(cli): Studio mounts the installed item's own file by the path add recorded

* fix(cli): Studio's install never hands back a file the project kept for its own edit
2026-09-27 09:06:15 +00:00
Miguel Ángel 950cccdac0 test: cli and studio-server tests remove the temp folders they create (#4566)
* test: cli and studio-server tests remove the temp folders they create

* test(cli): each temp folder is removed where it is made, even when a test fails

* test(cli): the one-liner helper relies on its project's own cleanup
2026-09-27 04:06:05 +00:00
Miguel Ángel d290c05b09 fix(cli): --yes in an unattended run no longer overrides a no saved by another process (#4547)
* fix(cli): --yes in an unattended run no longer overrides a no saved by another process

* fix(cli): every settings writer takes the lock, and --yes needs a real terminal to reverse a no

* fix(cli): the thin-results download offer also needs a real terminal

* fix(cli): one owner-checked settings lock, and an unreadable settings file never counts as a yes

* fix(cli): settings lock removals go through a reaper, and a non-boolean answer counts as a no

* fix(cli): a settings lock that cannot be released no longer fails a write that landed

* fix(cli): a leftover settings lock is never taken over, and doctor names it

* fix(cli): a background update that cannot be recorded is not launched again every run

* fix(cli): every on-device install re-reads the saved answer first, and an unsaved yes installs nothing

* fix(cli): the thin-results offer stops when a yes cannot be saved

* fix(cli): only the consent-checked step installs the on-device runtime, and a leftover lock warns once

* fix(cli): the leftover-lock warning follows the attended-terminal rule
2026-09-27 01:39:17 +00:00
Miguel Ángel 7c634426fd docs(cli): explain how catalog search works and which model it downloads (#4462)
* docs(cli): explain how catalog search works and which model it downloads

* docs(cli): name the model catalog search was chosen against

* docs(cli): keep only live rules in the word-search header comment

* docs(cli): correct the hosted vector size in the local vector build comment
2026-09-25 04:43:54 +00:00
Miguel Ángel 8dabc8af63 fix(registry): glass-shard-title renders when mounted, and 4 skipped blocks install again (#4423)
* fix(registry): glass-shard-title renders when mounted, and 4 skipped blocks install again

* fix(registry): glass-shard-title loads in Studio and ?play, and its docs payload fails loudly

* test(scripts): pin the glass payload's hdr guard for any quote style
2026-09-24 13:08:15 -04:00
Miguel Ángel 0aa3cb82aa fix(cli): add --vars refuses a value of the wrong type instead of writing it (#4369)
* fix(cli): add --vars refuses a value of the wrong type instead of writing it

* fix(cli): a refused --vars writes none of the item's dependencies either

* fix(cli): keep a file edited while add was still downloading the rest of the plan

* style(cli): shorten the --vars refusal comment
2026-09-23 19:29:16 -04:00
Miguel Ángel 05d09a7452 chore(registry): remove low-use catalog items and variant tails (#4274)
* chore(registry): remove low-use catalog items and variant tails

* fix(registry): remove stale references to deleted catalog items

* style: format catalog fixture

* fix(ci): use surviving catalog preview canary

* chore(ci): allow intentional catalog deletions

* fix(ci): scope fallow audit for catalog fixtures

* chore(catalog): regenerate local search vectors

* chore(catalog): update vector revision

* fix(catalog): validate registry set delta
2026-09-22 07:04:54 -04:00
Miguel Ángel d5e00c0e66 feat(registry): week in merges leaderboard block (#4264)
* feat(registry): add a week-in-merges leaderboard block

Authors race across the row at a speed set by their merge count, bouncing at both edges.

* chore(registry): cover week-in-merges in the catalog search index

* style(registry): format the week-in-merges block

Wrap the lines oxfmt rejects so the format check passes.

* fix(registry): regenerate the week-in-merges catalog payload

The format pass left the committed payload off the generator output.

* feat(registry): cap week-in-merges at twelve rows

The block races at most rows_max authors, highest merge count first, and writes one line for everyone left out. The default is 12, so a long list cannot shrink the rows.

* fix(registry): make week-in-merges speed follow merge count

A row now travels the leader's pace times its merges over the leader's count. Only a zero stays on the floor, so a near-tie no longer looks like last place.
2026-09-21 23:35:55 -04:00
Miguel Ángel 3091dcb787 feat(cli): route media search through media-use (#4230)
* feat(cli): move media-use engine into the CLI

* fix(cli): replace media-use compatibility asset

* fix(cli): replace moved media-use tree

* fix(cli): copy moved media-use files cleanly

* fix(cli): copy media-use engine files individually

* fix(cli): keep audio helper compatibility import

* fix(cli): clear copied skill destinations

* test(cli): preserve moved media-use test paths

* fix(cli): preserve media-use script compatibility imports

* fix(cli): preserve media-use telemetry shim

* fix(cli): resolve packaged media-use LUT assets

* fix(cli): resolve packaged bundled SFX assets

* chore(skills): regenerate media-use manifest

* fix(cli): preserve media-use parity coverage

* fix(cli): cover relocated media-use tests

* chore(skills): regenerate media-use manifest

* fix(cli): format bundled media provider

* fix(ci): run relocated media-use tests with node

* fix(ci): preserve fallow script exemptions

* fix(ci): scope media-use fallow exemption

* refactor(cli): trim media-use essay comments

* feat(cli): index media-use assets for catalog search

* feat(catalog): commit media-use search vectors

* chore(catalog): track media vector metadata

* fix(catalog): validate media vector row order

* fix(catalog): validate media artifact freshness

* fix(catalog): cover media metadata and credits

* fix(cli): reject malformed media rows

* fix(catalog): compare complete media rows

* feat(cli): route media search through media-use

* refactor(cli): trim relocated media-use comments

* fix(cli): make media search runtime-safe

* fix(cli): preserve sfx fallback diagnostics

* fix(cli): probe sfx fallback provider explicitly

* chore(skills): regenerate manifest

* refactor(cli): keep typed media ranker source

* fix(cli): align media search checks with local-first flow

* fix(cli): scope vector audit exceptions

* fix(cli): ship resolver registry beside skill

* fix(cli): keep media-use skill scripts standalone

* test(cli): cover local vector validation

* test(cli): correct media vector fixtures

* fix(cli): prefer exact media keys

* fix(cli): stem exact media keys

* test(cli): skip media lint for snippets without media

* fix(cli): align skills test discovery

Keep the CI skills job's bash discovery as the single source of truth so its empty-list guard covers the exact files Node executes. Make hasMediaVectors private because it has no external callers and needs no Fallow suppression.

* fix(cli): close media-use review gaps

Prioritize literal ids before stemmed ids and cover the collision alongside the plural regression. Add command seam tests for engine discovery, actionable failure, and argv passthrough. Add a deterministic parity check for standalone media-use copies, with documented allowlists for the standalone media-fetch shim and self-contained npx resolver. Keep the bash-discovered skills test list as the CI source of truth.

* fix(cli): lower parity check complexity

Split missing-copy and unallowlisted-drift checks so the parity guard keeps its behavior without a threshold CRAP score.

* fix(cli): close media-use review seams

Run the parity test in the skills job so skills-only edits exercise the same check. Export the verb mapping and inject command dispatch for direct coverage. Keep missing-copy diagnostics separate from drift reads.
2026-09-21 18:35:47 -04:00
Miguel Ángel e8fdfd0d65 fix(cli): install onnxruntime-node and google-genai on first use (#4156)
* fix(cli): install onnxruntime-node and google-genai on first use

Both carry deprecated transitive dependencies (boolean, node-domexception)
that showed up as npm warnings on every npx hyperframes install. They move
to devDependencies (types and the Intel-binding check still need them at
build time; the bundler externalizes them) and load through a new
on-demand loader that installs into the CLI's own cache dir the first
time a feature needs them, with one stderr line, no prompt, and a
manual-install command named in the error if the network is unavailable.

hyperframes doctor lists each as installed (with version) or
not-installed-yet.

* fix(cli): sweep stale install staging and bound network retries

* fix(cli): optional-package install only sweeps staging dirs whose process is dead

Staging dirs now carry a per-call suffix after the pid, and the sweep skips any dir
whose pid is alive, so concurrent first runs no longer delete each other's install.

The install error suggests the network only for network failures, and the manifest
path has one owner. install() takes an injectable npm runner so tests drive the real code.

* fix(cli): optional-package sweep also clears old-format staging dirs and tests a live foreign dir

The concurrent-callers test shared one pid, so it never exercised the sweep rule; a live foreign
staging dir test now does. Old-format dirs of a dead pid are swept, and a failed cleanup no
longer aborts a viable install.

* fix(cli): optional-package sweep names a stale dir it could not remove

A locked leftover is still skipped so the install proceeds, but the warning makes it visible.

* docs(cli): shorten the install docblock to the comment limit

Comment-only change.
2026-09-19 05:27:35 -04:00
752b552789 feat(registry): blur an element by marking it, not by calling a function (#4045)
attachMotionBlur(el, tl) has an ordering contract: it has to run after every
tween is defined and before the composition registers its timeline. Getting it
wrong makes the blur silently do nothing, which is the worst shape a failure can
take in a renderer nobody watches. An attribute has no order to get wrong.

data-hf-motion-blur on any element is now the contract. Empty means defaults;
anything else is a JSON object of the same options. The element finds its own
timeline: its nearest [data-composition-id] ancestor names the key, and the
snippet reads window.__timelines[key]. attachMotionBlur stays exported for a
target built after the page has settled.

claim(el) is the only thing that attaches, and two triggers call it. An accessor
on window.__timelines wraps the registry in a Proxy whose set trap sweeps on
every per-key registration, which is what attaches before the renderer captures
frame zero. A bounded poll, 32 ms by 250 ticks, reads the registry directly and
needs no cooperation from whoever wrote it. The poll is not belt and braces: a
registry reference captured before the snippet installs bypasses the trap
outright, and a composition that mounts its DOM after registering its timeline
has no marked element to sweep at trap time. It runs its whole budget every time
rather than stopping the first tick nothing is pending, because on an ordinary
page the top-level target is claimed on the registration write, so an early exit
kills the poll one tick in, before any late composition arrives.

Every refusal is by name, once, and never silent: a value that is not JSON, a
value that parses but is not an options object, an option name that is not one
of the four, an option value that is not a finite number, a target inside
another target, and a second timeline registered under one key. Parsing runs
before the timeline lookup, so a malformed value is reported as malformed rather
than as a composition that never registered.

Two properties of the primitive this change made load-bearing:

- Frame rate is resolved from the target's own composition root, not from the
  first root in the document. The document-wide lookup was survivable while
  every caller could pass fps explicitly; the declarative form has no per-call
  argument. Two compositions on one page at different rates now each get their
  own shutter window, and one call naming both says it cannot.
- An attach is all-or-nothing. The mark, the copies and the resize observer go
  in together and come back out together, so a throw partway through leaves the
  element untouched rather than marked blurred with copies in the page and no
  tracker driving them. The sweep is also bounded per element, because the write
  trap runs it inside the author's own registration statement.

Copies are stripped of data-hf-motion-blur, data-composition-id and data-fps,
on the clone and every descendant: the first would make a copy a target of its
own, and the other two would let a copy win the fps lookup, since copies are
inserted before the original.

skills/hyperframes-animation/references/motion-blur.md is the agent-facing half,
mirroring how colour grading is surfaced. It leads with when NOT to blur,
because the attribute makes blurring everything trivial and a smear on motion
that was never fast enough reads as a soft, cheap render.

Co-authored-by: Miguel Ángel <miguel.sierra@heygen.com>
Co-authored-by: Vance Ingalls <vance@heygen.com>
Co-authored-by: Claude Code <noreply@anthropic.com>
2026-09-18 22:55:15 -07:00
miga-heygenandMiguel Ángel 82f49262ec feat(registry): blur every transform driver, and ship the reference case as shutter-slam (#4038)
The motion-blur snippet integrated translation only. Its output stage was one
SVG filter per target whose copies were feOffset, and feOffset only translates,
so a beat that only scaled or only turned sampled a displacement of zero and
rendered sharp. No SVG filter primitive can give a copy its own affine
transform, so the sampling loop was never the problem: the output stage was.

Copies are now DOM clones of the element in an isolation:isolate group, each at
1/N opacity with mix-blend-mode plus-lighter, each carrying the element's
resolved transform at its sample time. plus-lighter adds premultiplied colour,
so N + 1 copies at 1/N sum to the average the shutter integral asks for; the
isolation is load-bearing rather than tidy, because without a transparent
backdrop of its own the first copy adds onto the page and blows a light
background out to white. Reading the resolved matrix instead of a list of named
GSAP properties is what makes every driver work at once: translation, scale,
rotation, 3D rotation and skew arrive in the same one value.

Measured against the After Effects export the shutter model was fitted to, per
beat, previous stage to this one: scale +0.62 dB, rotate X +0.43, rotate Y
+0.84, rotate Z +1.15, translate X -0.18, translate Y -0.04. The two small
translate regressions are inherent and not tunable: the filter offset one
rasterisation so every copy was pixel-identical, where each copy is now
rasterised independently at its own fractional transform with its own subpixel
antialiasing.

Three consequences of the new stage, each carrying its own fix:

- mix-blend-mode flattens preserve-3d, so a copy cannot inherit the parent's 3D
  context. Each copy carries the parent's perspective as its own first transform
  function instead.
- A copy is styled by nothing that selected the original, since dropping the id
  also drops every #id rule that gave it size, colour and font. Each copy
  carries its own resolved style inline, read once and replayed. Those values
  are px, so a container-relative element needs them again when its box changes:
  a ResizeObserver re-snapshots rather than paying a subtree walk every frame.
- A beat that moved and faded at once kept a full-strength smear behind a
  vanishing element. The element's opacity now rides the group; the 1/N weight
  stays on the copies.

The axis option is gone. It selected which of x and y to sample, a question the
resolved matrix does not have.

shutter-slam is the new component: one word through six single-property beats,
two translate slams, a scale punch, then a full turn about each axis, which
exercises every driver the model covers in one mount. Easing is power2.inOut
because it measured best, not because it was the default. A bezier fitted to the
reference's own bbox centres looks like the better curve and renders worse: a
blurred bbox centre is the shutter window's average rather than the instant, so
on an accelerating segment the fit measures its own blur.

Co-authored-by: Miguel Ángel <miguel.sierra@heygen.com>
2026-09-17 16:03:14 -04:00
miga-heygenandMiguel Ángel 9aa60a844f fix(registry): make motion-blur match After Effects shutter-based blur (#4008)
* fix(registry): make motion-blur match After Effects shutter-based blur

The motion-blur catalog snippet estimated speed from the x/y delta
between consecutive timeline updates and drew three one-sided ghost
copies with decaying alpha plus a Gaussian on top. After Effects does
temporal supersampling: each frame integrates the layer over a shutter
window (shutter angle, shutter phase, samples per frame), producing a
symmetric box smear of length speed x shutter time.

The snippet now does the same inside the page. After every timeline
update a deferred pass seeks the GSAP timeline to N sub-frame times
across the shutter window (events suppressed), reads each target's x/y,
restores the frame time, and drives an SVG filter that averages N
feOffset copies with equal weight (feComposite arithmetic, 1/N, sRGB).
Defaults follow AE: shutterAngle 180, shutterPhase -90, samplesPerFrame
16; fps comes from the render config, then the root data-fps, then 30.
Displacements are mapped through the inverse of the element's own
scale/rotation; percent x/y values are resolved to px; SVG targets use
their bbox; an existing inline filter is kept and the blur appended to
it. The pass runs in a microtask rather than inside the tracker tween's
onUpdate, because seeking GSAP from within its own render reads stale
tween state and suppresses the composition's callbacks on forward
seeks. The heuristic options blurScale/blurMax/stretchScale/stretchMax
are removed.

Reference: an AE-algorithm reconstruction of the same composition
rendered without the component at an effective 960 fps and averaged
over the 16 sub-frames of the 180 degree window per output frame.
Against it (1920x1080, 30 fps, solid + text translating linear then
power2.inOut): old mean PSNR 53.2 dB (min 26.7), SSIM 0.9930; new mean
PSNR 72.4 dB (min 52.6), SSIM 0.9998. The 10-90 edge ramp at 66.7
px/frame is 26.8 px in the reference, 27.5 px new, 14.5/6.3 px
(one-sided) old; at the ease peak 79.3 vs 79.5 vs 7.9/17.7 px, with
smear centres matching to 0.1 px.

Co-Authored-By: Miguel Ángel <miguel.sierra@heygen.com>

* fix(registry): drop render-config fps probe from motion-blur and refresh its docs

The snippet resolved fps by reconstructing the renderer's private seek
config key so the compiler's inline-script guard would not strip it. A
public catalog snippet must not depend on that key, so fps is now the
explicit option, then the root data-fps, then 30; renders with an fps
override pass fps explicitly. The catalog page's Usage, How it works and
Options sections still described the old velocity heuristic and are
rewritten for the shutter model. The 1/N weight keeps full precision.

Co-Authored-By: Miguel Ángel <miguel.sierra@heygen.com>

* fix(registry): match the After Effects reference export's shutter window

The shutter model landed on this branch was built against a reconstruction
of After Effects' algorithm, not against After Effects. A real 1920x1080 /
30 fps export of translating text disagrees with it on three points, all
measurable off the export itself.

The window is two frames wide, not half a frame. On the measured frame the
outermost duplicate on the trailing side sits 128 px back while the previous
frame is 125.8 px back, and the outermost leading duplicate sits 189 px ahead
while the next frame is 184.9 px ahead: the window ends land on the
neighbouring frames. Defaults become shutterAngle 720, shutterPhase -360.

Duplicates sit on the sub-interval boundaries, both ends included, not at
sub-interval centres. The staircase across a stroke steps every 15.6 px on a
125.8 px displacement and every 23.0 px on a 184.9 px one, i.e. an eighth of
a frame, with the run closing exactly on the neighbouring frame's position.
samplesPerFrame now means sub-intervals of the window, so there are N+1
copies at k/N rather than N copies at (k+0.5)/N.

The frame-time instance is drawn sharp over the smear. At full ink the core
bounding box is the whole word at every frame and total ink rises to 1.44x
the unblurred frame's, neither of which an average can produce. A final
feComposite over of SourceGraphic reproduces it. Each duplicate stays at
1/N: the staircase plateaus step by 0.063 +- 0.002 of the sharp text
intensity, flat, with no taper toward the window edges, and triangle
weighting scores 1.6 dB worse against the export.

Verified by replaying the export's translating section as a composition
whose sprite is the export's own rest-frame ink and whose eight legs follow
its own measured per-frame positions, rendered at 1920x1080 / 30 fps and
compared frame by frame over the text band. Mean PSNR against the export
rises from 24.19 dB (min 17.07) to 32.51 dB (min 27.57) across 72 frames,
and from 19.83 dB (min 17.07) to 32.03 dB (min 27.57) across the 26 frames
that actually move. Ghost extents match to a few px and the rendered
staircase plateaus match the export's to two decimal places. The export's
later scale, skew, flip and 3D rotation sections are out of reach here:
integrating anything but translation inside the shutter is engine work
tracked in issue #4010.

The filter region now clamps its bounds through zero. It was derived from
the duplicate offsets alone, which was correct while the output was nothing
but duplicates; the sharp instance draws at offset zero, so a window sitting
entirely on one side of the frame time - any shutterPhase above 0 or below
-shutterAngle - produced a region starting inside the element's own box and
cut the instance off. The defaults straddle the frame time and were never
affected, so the measured numbers above are unchanged.

Co-Authored-By: Miguel Ángel <miguel.sierra@heygen.com>

---------

Co-authored-by: Miguel Ángel <miguel.sierra@heygen.com>
2026-09-17 00:51:03 +00:00
renovate[bot]andJames eae4892ae8 chore(deps): update dependency vitest to v4 [security] (#3789)
* chore(deps): update dependency vitest to v4 [security]

* fix(test): preserve test behavior on Vitest 4

---------

Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
Co-authored-by: James <james.russo@heygen.com>
2026-09-10 13:55:52 -04:00
James Russo caa72de355 fix(cli): bound and validate registry installation (#3814)
* fix(cli): bound and validate registry installation

* fix(cli): preserve registry install compatibility
2026-09-09 12:41:48 -04:00
3a7fcd10e0 feat(registry): add 25 image carousel blocks (#3790)
* feat(registry): add 25 image carousel blocks (5 families × 5 variants)

Five carousel families, each with 5 style variants:

- Orbit (1–5): image cards on a spinning 3D Fibonacci sphere
- Path (1–5): cards following animated CSS motion paths
- Circle (1–5): circular carousel layouts
- Vision (1–5): Apple Vision-style spatial presentations
- Text Circle (1–5): circular carousels with text overlays

All blocks are 1920×1080 at 6s, with 12–24 configurable image slots.
Includes catalog preview thumbnails for each block.

Co-Authored-By: Jake Moran <jake.moran@heygen.com>

* refactor(registry): host carousel block images on the CDN

The registry is served straight out of this repository
(DEFAULT_REGISTRY_URL points at raw.githubusercontent.com), so every byte a
block ships is permanent history. The 25 carousel blocks added 421 JPEGs,
27.7 MiB in a checkout, and made the diff 472 files. Only 23 of those images
were distinct: the same 12-24 placeholders were copied into every block.

files[] entries gain an optional `url`. When set, the installer fetches the
bytes from there instead of joining the registry base. `path` does not change
and still says where the file lands relative to the item, so composition HTML,
target mirroring and `hyperframes add` behave exactly as before.

Keys are content-addressed, so the 396 manifest entries resolve to 23 objects,
and a changed image gets a new URL rather than a stale one cached behind
`immutable, max-age=31536000`.

The catalog preview renderer copies an item's directory and renders it, so it
needs the same materialisation step. Without it the preview draws every card
blank and reports success, which is worse than failing.

Also drops registry/catalog/, 25 hand-made thumbnails referenced by nothing;
catalog previews are rendered by CI and served from docs/images/catalog.

Verified: all 23 objects return 200 from the CDN with hashes matching their
keys; `hyperframes add carousel-orbit-1` against a local registry installs 24
real JPEGs; the preview render produces the album art, and produces blank
cards when the fetch step is removed.

* style(registry): format the carousel composition HTML

`oxfmt --check .` covers the whole tree, and these 25 files were never run
through it. The pre-commit hook only formats staged files, so nothing local
caught it.

* feat(catalog): publish the carousel blocks without republishing their images

These 25 blocks had no Catalog page. Every other item in the registry has one,
so they shipped invisible: installable by name, unfindable by browsing.

Generating them naively undid the change they were added by. The Catalog
payload copies an item's assets into docs/public/, which is tracked, so the 396
images this PR just removed came back as 43 MB one directory over — worse than
the 3 MB they started as, because each block got its own copy.

The copy exists because these compositions assemble `img.src` at run time out
of a variable value, so there is no `src="..."` in the markup for the payload's
asset scan to resolve. An unpredictable path can only be satisfied by serving
every file beside it, which is what `needsOwnDirectory` asks for.

An absolute URL needs no directory: the scan already skips any `https:`
reference. So for the payload path only, hosted files are left undownloaded and
the composition's variable defaults are rewritten to their URLs. The preview
renderer still downloads them, because it paints real frames and a missing file
is a blank card.

The explorer posts every value to the preview frame on mount, including
untouched ones, so the page's variable list carries the URLs too. Left as local
paths they would have overridden the payload's own defaults and asked the frame
for a file that was deliberately never published.

Result: 25 pages, 25 payloads, zero bytes of image added.

Verified: a spike item declaring no assets at all rendered its 24 covers from
the CDN, proving the variable-default path; payload generation for a carousel
block now writes no item directory and no shared asset; the preview render
still produces the album art. mint validate and mint broken-links pass on the
new pages. test:scripts is green.

* refactor(catalog): split the hosted-asset step out of prepareProjectDir

Two functions rather than one: finding the composition and rewriting its
variable defaults are separate jobs, and inlining the mode branch pushed
prepareProjectDir past the complexity gate it was already sitting on.

Behaviour is unchanged. Re-verified both paths after the split: the payload for
a carousel block still writes no item directory and no shared asset, and the
preview render still produces the album art.

* feat(catalog): give the carousels their own shelf

25 image carousels landed in Showcases and were 53% of it, so the scenes that
shelf exists for disappeared underneath them. That is the same shape the 24
editor themes made, and it gets the same fix they got.

Keyed on the first tag, which is this file's stated grouping rule, rather than
on the name. `screen-flow-carousel` leads with `product-demo` and stays on the
shelf that says what it is for; a future carousel that is not named
`carousel-*` still lands here.

Showcases 47 -> 22, Carousels 25, and no existing item changed shelf.

* fix(registry): centre the circle-5 carousel path in its composition

Its ring was centred at x=3832.6 in a 3840-wide composition, so it sat on the
right edge and most of it fell outside the frame. Only a few cards were ever
visible, cropped, with two thirds of the composition empty.

The exported path carried absolute coordinates from a layout that was never
recentred. Shifting the four vertices by (-1912.597, -4.340) puts the ring on
the composition centre. Handles are relative, so only the anchor points move
and the shape is unchanged.

carousel-text-circle-5 shares the identical path and had the identical fault.

The other three circle variants sit within 12% of centre, which reads as
authored placement rather than the same bug, so they are left alone.

* fix(catalog): rebuild the circle-5 payloads after recentring the path

The Catalog preview plays the payload, not the composition on disk, so
recentring the source changed nothing a reader sees. The payload still carried
the old vertices and the ring still hung off the right edge of the frame.

Verified the consumer this time, not just the producer: both payloads now
resolve to a path centre of x=1920. The other 23 rebuild byte-identical, so the
formatting pass did not reach them.

* feat(catalog): promote Carousels to its own section

It was a shelf inside Scenes & demos, which is where a scene type belongs by
kind but not by weight. At 25 items it is larger than Data & charts (17) and
Blocks (13), each of which is already a section holding a single shelf, so the
catalog's own precedent puts it one level up.

Pulling it out also takes the largest section in the catalog from 120 items to
95, which is the reason the shelf was added in the first place.

The two circle-5 pages change because their embedded source block carries the
recentred path; nothing else in them moved.

---------

Co-authored-by: Jake Moran <jake.moran@heygen.com>
Co-authored-by: Miguel Angel Simon Sierra <miguel.sierra@heygen.com>
2026-09-09 00:47:31 -04:00
James 71e8e92ae9 fix: document safe template default editing 2026-08-25 06:54:02 +00:00
James 1e11608d44 fix: expose promoted template media slots 2026-08-25 05:33:50 +00:00
James Russo dac8f9f912 feat: add promoted template edit contracts (#3407)
* feat: add promoted template edit contracts

* fix: address template contract review feedback
2026-08-21 18:39:17 -07:00
Miguel Ángel ea7c48f372 fix(add): make chosen variables actually take effect, in the CLI and the preview (#3316)
* fix(add): apply --vars to components, and explain a failed download

Customising an item on the catalog page, copying the printed command and
running it did nothing for a component. `--vars` was accepted, documented
and then dropped: buildSnippet put the values on a block's mount element
and returned a bare "paste from ..." comment for a component, so 221 of
the 375 catalog items silently ignored every value the page produced.

A component has no mount element to hang values on. It is markup pasted
into a host, and it resolves values through __hyperframes.getVariables(),
which merges the declared defaults of every [data-composition-variables]
element in the document with render-time overrides. So the component's
own declaration is the only place a chosen value can live and still be
there after the paste. `add --vars` now rewrites those defaults.

Blocks keep the mount attribute. Per-mount values are strictly better
where a mount exists: the file on disk stays byte-identical to the
registry's, so a later reinstall can still tell an edit from an update,
and two mounts of the same block can differ.

A value the item cannot accept is now refused rather than written. An
out-of-range number or an unlisted enum value falls back at runtime and
warns, so writing one would produce a file that renders exactly as if the
value had been ignored -- the failure this change exists to remove. Ids
the item never declared are reported too, instead of vanishing. Only the
requested item is rewritten; a dependency dragged in behind it never
declared these variables.

Separately, `Install failed: fetch failed` is now a sentence. Item FILES
are not cached (only manifests are), so a network blip surfaces as node's
bare message with no URL and no cause, immediately after the user copied
a command off a web page -- which reads as "the command was wrong" rather
than "the network was". It now names what failed, says it is usually
connectivity or a proxy rather than a bad command, and mentions
HTTPS_PROXY.

Also fixes the two transcribe tests that were failing before this branch.
They assert the whisper soft-skip path but never pinned the engine, and
`auto` picks Parakeet whenever parakeet-mlx is installed -- so on those
machines the test shelled out to a real ASR binary, failed with "Parakeet
did not produce output", and landed in the generic failure branch it
claims is never taken. Pinned to `engine: "whisper"`, plus an assertion
that the mocked transcribe actually ran, which is what stops the test
passing on a machine without Parakeet while testing nothing on one with
it. The file now runs in 18ms rather than 3.7s, because it no longer
launches a subprocess.

Test plan: 10 new tests for the rewrite (enum and range refusal, the
numeric-string coercion the catalog URL depends on since every query
value is a string, delimiter escaping, unparseable declarations) and 3
for the failure message. Full CLI suite: 2661 passed, ZERO failures.

Verified as a user, not just in unit tests: installed blur-in with the
exact reported command, confirmed the declaration carried 76 / accent /
center, pasted it into a composition and ran `check` -- which reported
canvas_overflow at 76px, which only happens if the baked size is really
in effect. Bad values warn and are refused; blocks still emit
data-variable-values.

* fix(player): load the runtime before the body, not after

Customising a component on a catalog page did nothing to the preview.
badge-pop with count 10 and a green accent rendered 3, in red.

The probe injects the runtime by appending a script to an already loaded
document, and only once it has a reason to: a nested composition, or five
polls with a timeline present. A component has neither. It is markup
pasted into a composition, and it reads its values in an inline IIFE that
runs while the body is parsing:

    var vars = window.__hyperframes && window.__hyperframes.getVariables
      ? window.__hyperframes.getVariables() : {};

With the runtime arriving afterwards that guard always took the empty
branch, so the component used the defaults hardcoded in its own script
and every chosen value was dropped. The values were never the problem:
the preview sets window.__hfVariables correctly, and nothing was there to
read it.

prepareSrcdocForElement now puts the same runtime URL in the document's
head before the srcdoc is set. A classic external script in head is
parser-blocking, so it runs before body scripts without changing what
gets loaded or adding a dependency the player did not already have. A CLI
render never had this bug because the engine already orders it this way.

Skipped when the page carries the runtime already, so a CLI-rendered page
(which inlines it) does not get a second copy re-initialising the runtime
underneath a live composition. The probe's late injection stays for the
src= path, where there is no srcdoc to prepare. The runtime URL moved to
its own module so the two injection points cannot drift apart.

Test plan: 8 new tests for the injection (ordering against the reading
script, head placement, both no-op guards, missing head/body, attributes
on the head tag). Three srcdoc tests asserted byte-identical forwarding
and now assert what they were actually protecting -- that the composition
arrives intact -- plus the new runtime guarantee. player 338 passed,
studio 4249 passed.

Verified end to end against the real runtime and a real registry
component, asking for size 96 / accent / right:
  before  52px, rgb(243,243,243), flex-start, runtime absent
  after   96px, rgb(60,230,172),  flex-end,   runtime present
rgb(60,230,172) is #3ce6ac, the accent green. That is the reported bug
before, and the chosen values after.

* fix(add): name the registry and the real reason an install failed, and retry

`Install failed: fetch failed` was two words that describe every network
problem equally badly. Three things were missing, and each of them was
the whole answer in a different case.

The URL. undici throws with no URL attached, so a project that points
`registry` at a private host in hyperframes.json got a message that
looked like the public registry had failed. Naming the URL is the entire
diagnosis there.

The cause. undici buries the real reason one or two levels down in
`cause`, and it was being dropped. The reported failure turned out to be
`self-signed certificate in certificate chain`: a private registry whose
certificate node refuses and curl accepts, which is why the host looked
healthy from a terminal. That sentence tells the reader which knob to
turn; `fetch failed` sends them to check a connection that is working.

The retry. Item files are the one uncached path -- manifests fall back to
a stale copy, but every install downloads its files fresh -- so a single
blip killed the whole command. Now two extra attempts with short backoff,
and deliberately NOT for TLS failures: a self-signed certificate fails
identically every time, so retrying it only makes the user wait three
times as long for the same message.

Also retypes the declaration reader. It modelled variables as a local
interface of six `unknown` fields and re-checked each one at every use.
Core already owns this shape as a discriminated union and exports
`isCompositionVariable`, the same predicate `parseCompositionVariables`
filters with, so the union is used directly and the duplicate type is
gone. A declaration the schema rejects now leaves the file untouched
rather than being partially rewritten from guesses.

Test plan: 4 retry and URL tests, 5 cause-chain tests, and the add-side
tests now cover the custom-registry hint and its absence on the default
registry. The variableDefaults fixtures gained the `label` the schema
actually requires; without it they were not valid declarations, which the
stricter reader caught. CLI suite 2671 passed, zero failures.

Verified with the BUILT dist rather than the source, in the reporter's
own project directory. The failure now reads:

  File fetch failed: https://<host>/registry/components/blur-in/blur-in.html
    - fetch failed (self-signed certificate in certificate chain
      [SELF_SIGNED_CERT_IN_CHAIN])

and once the project points back at the public registry the original
command succeeds with `variables applied: size, tone, align`.

* fix(registry): name the registry on the not-found path too

The item-file failure now names the host it could not reach, but the
sibling path did not. A project whose registry is unreachable at the
MANIFEST stage got `Item "blur-in" not found - registry unreachable or
empty`, which reads as the public catalog having lost the item and sends
the reader to search a registry that never saw the request.

Same fix, same reason, applied where the other three call sites live so
one of them cannot stay behind: the message names the host and says it
came from this project's hyperframes.json, and only when it is not the
public registry, so the common case stays short.

Test plan: 3 tests covering the private-registry hint and its absence on
the default registry and on no registry at all. CLI suite 2674 passed,
zero failures. Verified with the built dist against a host with a bad
certificate:

  Item "blur-in" not found - registry unreachable or empty. Contacted
  https://self-signed.badssl.com/registry, set by this project's
  hyperframes.json, not the public registry.

* fix(catalog): reconcile the two spellings of a compound word

`countdown` returned exactly one item, the only thing tagged with that
spelling. `count down timer` returned sixteen, and that one was in none
of them. The tokenizer splits on word boundaries, so the two spellings of
a single idea produced disjoint sets, and whichever phrasing an author
happened to type decided which half of the answer they saw. Neither half
was the whole answer: the one-word spelling hid count-up and
decline-chart, which are the two things you would actually build with.

Both directions now, each gated on the catalog's own vocabulary so this
can only add signal. A query token is split when both halves are words
the catalog uses, and adjacent tokens are joined when the compound is.
A word in neither form, like `timer` which appears in no item, is left
alone: this widens phrasing, it does not invent matches.

Everything inferred this way carries a fraction of a real token's weight.
That is the part worth keeping honest, because the first version relied
on the halves being statistically common in a 375-item catalog, which is
not the same as making them count for less. In a small corpus that
version let `type` matching the name of `type-match-cut` outrank
`typewriter` matching the name of `typewriter`: searching a word returned
something that merely contained half of it. Two tests written against
that real failure caught it.

All spellings now return the same 17 items, and each still ranks its own
exact match first: `countdown` leads with yt-circle-pointer, `count down`
leads with the two-word items, and count-up and decline-chart appear in
both.

Test plan: 6 new tests covering both directions, the identical-set
property that was the actual defect, exact-match precedence, an unknown
word left alone, and the typewriter case. Eval set unchanged at 33/39
top-1 and 39/39 top-3, so no query regressed. CLI suite 2680 passed.
2026-08-17 20:31:37 -04:00
Miguel Ángel 6b17c24f98 fix(catalog): rank on where a word appears and how rare it is (#3312)
* fix(catalog): rank on where a word appears and how rare it is

Word search returned the right move in the top three for 87% of a
39-query eval set built from real catalog intents. Three defects, all in
the same 75-line scorer, and all found by running the queries rather than
by reading the code.

A token matching an item's NAME counted exactly as much as one buried in
a description. Searching "typewriter effect on a title" ranked the item
literally called `typewriter` seventh, behind entries that merely mention
typing. Name and title now carry three times the weight: an author who
types a move's name is giving the strongest signal available and it was
being averaged away.

Plurals shared no vocabulary with the singular. "a stat that counts up
and then pulses once" matched nothing in a description reading "lands
with a restrained scale pulse", because `counts` is not `count`. Adding
detail to a query made results strictly worse, which is the opposite of
what a search should do. Plurals now fold, and only plurals: Porter would
fold `counter` to `count` and `values` to `valu`, merging moves that mean
different things.

Field weighting alone made one case worse, which is why inverse document
frequency is here too. "reveal a headline one line at a time" put every
item merely NAMED `*-reveal` on top, because one strong hit on the
catalog's most common word outscored several weak hits on the words that
actually narrowed it down. Rarity now scales each term.

Separately: a query in a script this ranker cannot index no longer
reports itself as an empty catalog. Tokenising on [a-z]+ leaves nothing
of a Japanese query, and returning "no items match" told the author the
catalog lacked a move it may well have, then invited them to file a gap
report about it. That case now says what actually happened and withholds
the gap prompt, since nothing was searched.

Measured on the same 39 queries, before and after:
  top-1  31/39 (79%) -> 33/39 (85%)
  top-3  34/39 (87%) -> 39/39 (100%)

Test plan: 13 new tests, each a real failing query reduced to the
smallest fixture that still reproduces it. Existing tests migrated to the
fields API (two callers total). Full CLI suite 2643 passed, 2
pre-existing transcribe failures unchanged. Verified against the real
CLI: "typewriter effect on a title" now returns typewriter first, and
"chat conversation between a user and an assistant" returns chat-message,
chat-thread, ai-chat-reveal instead of transitions-blur.

* docs(skills): say to query the catalog in English

The runtime message added alongside this explains an unsearchable query
after the fact. Saying it up front is cheaper: an agent that never writes
the query in Japanese never sees the error, never wastes the turn, and
never files a gap report about a component that exists.

Worth stating rather than assuming, because the mistake is a reasonable
one. On a Japanese or Chinese project the brief, the narration and the
captions are all in that language and the query naturally follows. The
rule is that the query language and the video language are unrelated:
describe the move in English, write the on-screen copy in whatever the
video needs.

Both skills that own `catalog --query` carry it, and those are the only
two that mention the command at all.

* fix(catalog): fail a non-English query instead of returning nothing

The message explaining an unsearchable query went to stdout and the
command exited 0. An agent that checks the exit code, which is most of
them, read that as "searched successfully, the catalog has nothing" and
went off to hand-author a move that is sitting in the registry. The
explanation only helped a human who happened to be reading the terminal.

It is bad input, not an empty shelf, so it now behaves like one: the
guidance goes to stderr and the command exits 1, matching what an invalid
--type already does. A genuine empty result, where the query parsed fine
and the catalog simply has nothing, still exits 0 -- that distinction is
the whole point, and both halves are pinned by tests.

The wording now also says what to do rather than only what happened:
search in English, and let the on-screen copy of the video stay in
whatever language it needs. That was the part agents were getting wrong,
since a Japanese project makes a Japanese query feel natural.

Test plan: 3 new tests covering the exit code, the wording, and the
genuine-empty case that must stay at 0. Also asserts the gap-report line
is absent, since nothing was searched and a report there is noise in the
one signal that tells us what to build. catalog.test.ts 32 passed;
commands + registry suites 887 passed with the 2 pre-existing transcribe
failures unchanged. Verified against the real CLI: a CJK query exits 1, a
genuine miss exits 0.
2026-08-17 17:20:22 -04:00
Miguel Ángel 37f8c48449 fix(catalog): survive an unreachable registry, and ask for the gap (#3299)
Serve an expired registry cache when revalidation fails, so one timeout against the registry host no longer reports the whole catalog as unreachable while a usable copy sits on disk.

Hand back the gap-report command at the moment a search comes back wrong: catalog --query prints it pre-filled on both tiers, and every --json search envelope carries it as report_gap. Report on either tier, since the on-device tier needs a consented download and every gap reported to date came from the word tier.

Document the gap channel in the registry skill, which owns hyperframes catalog and never mentioned it, and name the CLI commands no skill did.
2026-08-17 14:55:17 -04:00
Miguel Ángel 532caf7aa2 chore(catalog): remove internal source markers 2026-08-13 21:10:12 -04:00
Miguel Ángel 0c33b2dc7a feat(cli): keep your edits when you reinstall a catalog item (#3193)
* feat(cli): keep your edits when you reinstall a catalog item

Running add again overwrote whatever was on disk, so a project that had tuned
an installed block lost that work without being asked or told.

The installer now records a hash of each file as it installs it, and compares
before writing. A file that still matches is replaced as before; one that does
not is left alone and reported. A file we have no record of counts as changed,
which covers both a project that wrote the file itself and one that installed
before the record existed.

--force restores the old behaviour for when you do want the registry's version.

* test(cli): cover the dependency plan install path
2026-08-10 19:11:16 -04:00