* fix(skills): reverse on Free opens the upgrade page for the person
* docs(skills): reverse on Free opens the upgrade page with a quoted link
* fix(cli): reverse on Free runs billing subscribe credited to the run
* fix(cli): trim the reverse skill and drop a doc comment for lint budgets
* fix(cli): drop --page-motion-blur from cloud render
The flag sent page_motion_blur to the hosted render API, which never
accepted that field and rejects unknown fields. Motion blur is set in
the composition with data-motion-blur, and the runtime turns it off on
its own when it cannot render correctly, so the cloud flag is removed
with its request plumbing, test and docs row.
Local render keeps its own --page-motion-blur flag.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore(ci): allow deleting the cloud render page motion blur test
The deletion guard blocks removed files unless they are listed. The
test file only covered the cloud render --page-motion-blur flag that
this branch removes.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): prefer OAuth session over HEYGEN_API_KEY
A signed-in `hyperframes auth login` session now wins over an API key in
the environment, so a paid plan on the login is used even when the shell
also exports HEYGEN_API_KEY.
New order: HEYGEN_ACCESS_TOKEN (host app), the stored login (refreshed
when expired), HEYGEN_API_KEY, HYPERFRAMES_API_KEY, the saved API key.
The env key still comes first when HEYGEN_API_URL or HEYGEN_API_BASE
names a host app's gateway, and takes over for the rest of the process
when the login cannot be refreshed. The media-use audio helper follows
the same order.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): fall back to the env key only when HeyGen refuses the refresh
A 503, a network error or an overlapping refresh no longer switches the
process to HEYGEN_API_KEY. The fallback now needs a refused grant or an
expired login with no refresh token. A login another process stored since
is re-read first, and refreshes that overlap in one process share one
request. The media-use helper no longer treats HeyGen's own API base as a
gateway.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): keep the login when the refresh reply can't be read
Only a 400/401 from the token endpoint marks the grant as refused and
switches to the env key. A 200 whose body can't be parsed now throws and
leaves the login in place, so the next call refreshes again.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* docs(changelog): weekly digest 2026-10-26-2026-11-02
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* docs(changelog): embed the Oct 26-Nov 2 weekly video
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* docs(changelog): tighten two claims from review
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): audio --read organises the reading by stem
Measures where each stem is heard, leaves the drums' recurring hits to the
music while it plays, names the stem of each remaining effect, writes a
voice-free mix next to the stems, and describes each audible stem and the
whole recording with Gemini.
* feat(cli): audio --read reads stems as vocals, drums, bass and melodic
Piano, guitar and other are mixed into melodic.wav next to the stems; levels,
spans, descriptions, effects and pieces are read per group, and a group heard
for under two seconds or 5% of the recording is left out of the reading.
* feat(cli): audio --read describes the six stems and writes a BGM prompt from them
Stems are read one by one again; a stem heard under two seconds or 5% of the
recording is still left out. Each stem's description gives the times of its
changes and names the effects it hears. The overall and the BGM prompt are
written from those descriptions and the measured spans, not by listening to
the mix again; the BGM prompt orders the arrangement without seconds.
* fix(cli): audio --read writes no BGM prompt for a recording without music
Foley and effects in a rhythm are not music; the reading says so instead of a
prompt that would make a remix generate a track the reference never had.
* fix(cli): audio --read prompts name no example sounds
* fix(cli): describe audio through HeyGen and keep effects measured
* docs(cli): document measured effect events without sound kinds
* fix(cli): clear audio description review checks
Use portable stem path assertions and simplify response validation.
Keep auth-derived URLs out of cache hashes and trim redundant comments.
— Jerrai
---------
Co-authored-by: Xuanru Li <157947275+xuanruli@users.noreply.github.com>
When /v1/hyperframes/reverse/start answers gate "block" for plan "free", `reverse start`
saves the session, then exits non-zero. The error and its `fix` carry the upgrade link that
credits the purchase to this run: the server's `upgrade_url`, else the plans page with
`feature=reverse&rs=<session>`. Pro, unknown plans, `record`, and every error still go on.
The reverse-any-video skill tells the agent to pass that exact link on and stop.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat(cli): reverse start notes a reverse run with HeyGen and cloud render carries its session
`npx hyperframes reverse start <video> [--dir <project>]` probes the video (the same JSON
as `measure probe`), then, when signed in, calls POST /v1/hyperframes/reverse/start and saves
the returned session id in the project's hyperframes.json. It never stops the run: signed
out, offline, a 404 from a server without the route, or a malformed answer all continue
with the probe and no session. `cloud render` sends `reverse_session_id` only when the
project holds one, which only a successful /reverse/start writes.
`init` now treats a folder holding only hyperframes.json as empty, so `reverse start --dir x`
followed by `init x` works and keeps the session. The reverse-any-video skill opens with
`reverse start` instead of `auth status` + `measure probe`, with a fallback for older CLIs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore(cli): keep the comment share of touched files from rising
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): reverse start keeps unknown-plan sessions, prints the attributed upgrade link, takes only rvs ids
- A `plan: "unknown"` answer is a valid session: saved, `recorded: true`, printed as
`plan: null`; it never blocks and never falls back to the client-side plan read.
- On Free, `reverse.upgrade_url` is printed: the server's link, else the plans page with
`feature=reverse&rs=<session>` (the fields checkout reads). The skill passes that exact
link on instead of `billing subscribe`.
- Session ids must match the render endpoint's pattern (`rvs_<lowercase uuid>`). A saved
value that does not is ignored: not sent on `cloud render`, and `reverse start` opens a
fresh session.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* docs(skills): name no CLI version for reverse start's fallback
0.9.4 shipped without `reverse start` and skills go live from main before the CLI release,
so the fallback is described as "a CLI without it" instead of a version number.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat(producer): report browser-computed reference backgrounds
Preserve one initial root/page reference across local, streamed and distributed
render results without guessing from film pixels.
— Jerrai
* fix(producer): preserve plan compatibility and optional inspection
Keep reference metadata outside the v2 manifest hash and read it from the
verified execution-plan artifact. Static planning uses a non-gating page
inspection instead of requiring render-worker capabilities.
— Jerrai
* fix(producer): poll readiness independently of animation frames
Use timed polling while the render clock parks animation frames. Cover readiness
that changes after the first check with a short real-browser regression.
— Jerrai
* fix(producer): budget measured reference metadata runtime growth
Producer adds 2,228 compressed bytes over current main. Tests, fixtures and source
maps are already absent from its dist-only tarball. Allow 7,766 bytes of headroom
above the verified 837,234-byte payload while retaining the unpacked limit.
— Jerrai
* feat(skills): publish reverse-any-video, render it at high quality, and stop on a stuck shot
From Wanbi's Pro test of reverse-any-video (11-01, checked against 0.9.2):
- A standard install never found the skill: it was internal, and /hyperframes did not route to
it. It is now an on-demand skill on every catalog surface, and the routing line names it.
- The cloud render in SKILL.md had no --quality high, so the final rendered at standard.
- A re-shoot that stayed pending had no rule; the run waited four hours. people.md now runs the
command once more after generate's 30-minute stop, then treats the shot as failed.
* feat(cli): generate takes a language for --voice, so a Pro sign-in gets a HeyGen voice
Wanbi's Pro run fell back to Kokoro: --voice needed an id, and the only way to list ids was the
separate heygen CLI and its own login. /v3/voices answers on the HyperFrames sign-in, so:
- --voice <language> picks HeyGen's first public voice in it and names it in the output.
- A missing --voice lists six English voices to pick from instead of a bare error.
- A voice id still goes straight through, with no extra call.
* fix(cli): a voice picked by language stays with its request across re-runs
Rames: `--voice Chinese` re-picked HeyGen's first voice on every run, and the voice is part of the
body and the idempotency key. If the catalog reordered while a job was pending, the same command
could start a second paid job instead of resuming the first.
The pick is now bound to the request (every field, plus the account scope a job bills) in the
project's heygen-jobs.jsonl, or the CLI config dir outside a project, and a re-run reuses it.
Regression: a pending timeout, a reordered catalog, the same command, for voice and avatar; the
second POST keeps the voice, body and key.
* fix(cli): a voice pick waits for a known account instead of binding to none
Rames: when /v3/users/me failed, senderScope() was null and the pick was bound to an empty
scope; once the account came back, the same command missed that binding and could pick another
voice and paid key. pickVoice now stops before picking or posting while the account is unknown,
and never reads or writes a blank-account binding.
Regressions, both directions: unknown then known refuses with no POST, then picks once; known
then unknown refuses the re-run with no second POST, then resumes with the same body and key.
Also README: a non-interactive install gets all 8 skills, not 7.
Add #6973 and #6978, which merged ahead of the release PR and shipped in
v0.9.0 without an entry. Correct the PLAN_TOO_LARGE upgrade note: the
plan-size cap only applied to Plan v1, and Plan v2 never raises it.
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: cache paths follow HYPERFRAMES_CACHE_DIR
Every place HyperFrames keeps downloaded or built files now takes its
folder from one helper, hyperframesCacheDir() in @hyperframes/core. The
default stays ~/.cache/hyperframes; HYPERFRAMES_CACHE_DIR moves it.
* fix: register the cache-dir subpath for bun and node, and build inspector from core source
* test: the cache-dir override test uses a folder that is absolute on every platform
* fix: the opencv cache follows HYPERFRAMES_CACHE_DIR
* fix(cli): npm installs run in their staging dir, so no cache path reaches a command line
* fix(cli): whisper.cpp is cloned inside its build dir, so no cache path is a git argument
* test(cli): the whisper clone test finds no system whisper on Windows either
* fix(core): accept 65-point .cube LUTs, the size grading apps export
* fix(core): shrink LUTs wider than the GPU texture limit instead of rendering black
* fix(cli): keep browser sign-in open through checkout
The loopback listener timed out after 5 minutes, but plan checkout runs
inside the OAuth flow. A slower purchase ended on ERR_CONNECTION_REFUSED.
Raise the default to 60 minutes, matching pacific's plan-checkout handoff.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(cli): show the sign-in limit and how to recover
The waiting line now states the 60-minute limit and Ctrl+C. After a
timed-out browser login the CLI says to run auth login again.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(cli): drop unverified purchase claim, harden timer test, document limit
Review fixes: the timeout hint no longer says purchases stay on the account
(the CLI cannot observe billing); the no-timer-left test now asserts before
any explicit close() and uses node:http so undici timers cannot pollute the
count; docs mention the 60-minute window and Ctrl+C.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(cli): close the sign-in listener on every path and record Ctrl+C
Ctrl+C is now the advertised exit, but events only flush on exit, which a
default SIGINT skips, so an abandoned login vanished from telemetry. Record
it as aborted and exit 130 through requestCliExit. Close the listener on any
throw, so a failure before the wait can't hold the process for the full
window. Derive the waiting line and timeout message from the listener's own
limit, and name the constant for why it is long, which clears the comment
ratchet.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(cli): billing and Desktop handoff read their hosts from settings
* fix(cli): OAuth client and API hosts read from settings
* test(cli): host and client settings tests run the same on any machine
* chore(cli): format and refresh the skills manifest
* chore(cli): trim a comment
* chore(cli): keep the stylesheet as it was
* fix(core): motion blur preview no longer slows down later in the timeline
With data-motion-blur-events="on", which Studio's Subtle and Strong write,
both blur tiers ran timeline callbacks at every per-frame pose read. Each
frame's reads step back from half a shutter ahead to the frame, so a
callback that keeps state and replays from zero on a backward seek (a
ripple simulation, for example) redid its whole history most frames, and
the cost grew with the playhead.
The mover scan still runs callbacks, so motion set from a render(t)
callback is still found. With events on, it also scans once without them,
and the per-frame reads run callbacks only when the two scans disagree.
The eventful scan runs first, so tweens take their start values after the
callbacks before them and a one-off call fires as often as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(core): keep the motion blur comments to the invariant
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(core): motion blur reads run callbacks when a callback moves any box
Comparing which elements each scan finds missed a callback that pans a
mover's ancestor while a tween also changes that ancestor: both scans
found the same movers and the same changing elements, so the reads left
the pan out. The scans now compare every changing element's box at each
step, and the reads run callbacks when any step differs.
Raises @hyperframes/core's unpacked size budget to the measured size plus
the usual 5%: main already sat 0.98% over the old budget, and this change
adds 1,232 bytes, which tipped it past the 1% tolerance.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(core): silent motion blur reads land back on the frame as its own seek left it
After silent pose reads, the restore seek back to the frame left GSAP in a
different state than the frame's own eventful seek did:
- A step exactly on the frame (a .call() or set) was re-armed, so the next
frame fired it again. Calls fired 2-5 times per crossing, and a toggle
could end inverted.
- In renders, which keep no GSAP structure, landing back on the frame from
later skipped same-time steps' authored order, so the frame showed the
first of two sets instead of the last.
Events-off blur always read silently and had both. This PR moves events-on
roots without callback-driven motion (the Studio default) onto that path,
so both are fixed here.
The host's new holdFrame snapshots every landing mark on the root and scene
timelines before the reads, and its restore seeks back silently, redraws in
authored order where there is no structure, and puts the marks back. Each
call now fires once per frame crossing, as without blur, and same-time sets
land as authored.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(core): a held render frame draws its scenes first, so the root lands last
holdFrame redrew the root and then the scenes, the reverse of the frame's
own seek. When a scene and the root tween one property, the frame landed
on the scene's value instead of the root's.
The hold and landing move into the GSAP adapter as holdFrameMarks, which
takes the scenes and the root separately, so the order lives in one place
a real-GSAP test can pin.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli,skills): say a step's credits are its total, and keep cost out of retries
An out-of-credits refusal named what the step needs, and an agent read
that as the shortfall. It now says the number is the step's total.
After its last try, an agent told the person what was spent and what
another try would cost. The HeyGen reference now says not to.
* fix(cli): plan avatar credits from the HyperFrames price rows
An avatar's credits came from its old API price, about five times what
the plan charges, so --max-credits refused calls that fit and a top-up
asked for too much. They now use the Avatar IV photo row plus the
general avatar video row, 950 credits a minute.
* fix(cli): pin the avatar price test and keep pricing under the comment share
* fix(cli): plan avatar seconds on the Avatar IV row alone
The general avatar row is only checked at admission and billed on
seconds left after Avatar IV, so adding it to every second made a 15 s
script plan at 238 instead of 200 and a 220 limit refuse a call that fits.
* fix(cli): the credits link opens the web plan page, which is live
* test(cli): pin the production credits link too
* chore(skills): regenerate the skills manifest after merging main
* docs(cli): keep the person measure's doc comments to four lines
* docs(cli): keep the measure doc comments to four lines
* docs(cli): keep the measure comment lines to 120 characters
---------
Co-authored-by: Miguel Angel Simon Sierra <miguel.sierra@heygen.com>
Make a video, Studio app and Developers move into a Docs menu (each keeps its
own sidebar), with Free lessons, Community and the app changelog under them;
Catalog stays a tab. The top-bar copy of Ask Assistant goes (Ask stays in
search and the page's bottom bar); the logo opens the homepage; the button
reads Download.
* docs(catalog): put each item page's title, description and Docs text in the server HTML
Mintlify renders snippet components in the browser only, so everything inside
CatalogDetail (title, standfirst, the Docs tab) was missing from the HTML search
engines read: 448 catalog pages served Googlebot no body text. The generator now
writes the H1 and standfirst as page markup above the component and the Docs tab
as page markup after it, shown with that tab (docs/custom.css). The component no
longer draws its own title. Pages patched structurally only; registry drift on 14
pages is left for a regeneration of its own.
* docs(catalog): keep the title's top spacing the component gave it
* fix(docs): catalog titles and descriptions keep markdown markers literal
MDX reads text inside the static <h1> and <p> as inline markdown, so @_blue_sweater_
rendered as italic blue_sweater. mdxText now writes backslash, backtick, * _ [ ] ~ as
character references too, with a unit test; the four affected pages are re-escaped.
* feat(lint): warn when a url() in a linked stylesheet does not resolve beside it
* fix(lint): a percent-encoded url() in a linked stylesheet resolves like the browser does
* feat(cli): render --start and --end render only that span of the film as its own MP4
The span reuses the chunked render path, so frames are captured at their absolute times and sub-composition timing matches a full render.
* chore(producer): shorten the renderSpan comment
* fix(producer): a cancelled span render stops, and its real-render test runs in the integration lane
renderSpan checks the abort signal before each chunk part and before the stitch. The real-render tests move to their own file listed in the integration lane.
* chore(producer): put the renderSpan doc comment back above renderSpan
* revert: feat(skills): reverse any video (#6754)
Keep the reverse-any-video skill and its CLI billing commands out of main until
web subscriptions are live (/pricing and /account/plan return 404 today).
Re-land together with #6833 once those pages ship.
This reverts commit cb2389cf35.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore(scripts): allow the #6754 revert deletions
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Miguel Ángel