* fix(cli): keep whisper CJK tokens as separate timed words
whisper.cpp marks word boundaries with a leading space, but Chinese and
Japanese tokens never carry one, so parseWhisperCpp merged every CJK token
into the previous word, across segments too. A 9.5-minute Mandarin episode
came out as two "words" of about 280 s each, leaving captions and
sentence-building nothing to split.
A token now starts its own word when it crosses a CJK boundary, full-width
punctuation still attaches to the word before it, and cues end at 。！？.
* fix(ci): capture package name in release-age exclusion
* ci(skills): hide retired skills on ClawHub once the sync has run
clawhub sync stops publishing a retired skill but leaves its last version
listed. A new step inspects each RETIRED_SKILLS name under heygen-com and runs
`clawhub hide --yes` with its retirement notice for the ones still listed. A
hidden skill reads as not found, so later runs hide nothing; a dry run only
names the skills still listed.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* ci(skills): import the retirement manifest at the top of the unlist script
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
#5971 and #5990-#5992 added htmlInCanvasDefects.ts and its package export
after #5969 last raised this budget, taking core to 3935667 bytes unpacked.
main's Build has failed on that since 4371981b7.
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Blur that turns itself off is silent in the output, so an author who asked for
data-motion-blur="on" cannot tell. Have hyperframes check read the runtime's
decision through probe() and the reason codes from the previous change.
check reports motion_blur_off as a warning with a fix hint for a root that
asked for blur and got none (an iframe in the root gets its own hint),
motion_blur_no_movers when nothing moves enough to blur, and
motion_blur_unverified when the check browser has no html-in-canvas to decide
with. The console echo of the same message is dropped so a finding is not
shown twice.
check asks the runtime which root wants blur instead of reading attributes
itself, so data-root="true" is honoured, and after probing it suspends blur
painting for the rest of the session: its audits read plain DOM, which removes
the roughly 10x slowdown of auditing a wrapped composition. It never sets the
page's off switch, which would change the probe's answer.
Tests drive a finding from the page to the check output, with its fix and
without its echo.
Folded review fixes: R9 (check hint), R12, R13 (check half), T6.
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
Tooling and logs need more than "off". Give the runtime's report a reason
code for every way blur can be off (renderer-off, unsupported-browser, each
html-in-canvas defect, capture-failures), and two calls for tooling:
probe() runs the same scan a render would, without painting, and suspend()
stops all painting for the rest of a session without changing the reported
decision.
The engine prints one line per session from what the page reports:
`[motion-blur] on reasonCode=null movers=N failedFrames=K`. The first-frame
line cannot know about later failures, so the session end prints the same
line with "(session end)" when any frame fell back to plain DOM; a clean
session stays quiet. A render where every capture failed no longer reads as
"on" with nothing else.
Add the first test that reads blurred pixels: real Chrome with
html-in-canvas, the engine's frame order (seek, wait, composite, screenshot).
It checks the trail along the motion line, that what does not move matches
the sharp frame, that a box centered with translate is blurred where it
really is, and that a flipped backface card renders the plain frame. It
skips when Chrome or html-in-canvas is missing.
Folded review fixes: R13 (runtime half), R17 (engine half), T1 (the guards
that need reason codes), T3 (log-line and pixel tests).
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
Decide the html-in-canvas defects once for the whole timeline instead of at
the frame being drawn. The mover scan already seeks the timeline, so the
defect scan samples along it and a defect that only appears mid-timeline (a
rotated root, a flipped card) turns blur off from frame 0. Defects now block
on every Chrome version, so a render matches the preview it came from, and
GSAP's identity transform on the root no longer counts as one.
A per-frame check stays as a backstop for what scripts add later, and it
costs only that frame: a defect the scan missed (a short root shake, a
backface flip between samples) renders that frame sharp and logs once, and
blur resumes when the defect is gone. A block raised by the scan is lifted
when its cause is removed or the timeline changes; a capture failure stays
for the life of the instance.
The candidates a scan checks include faces whose backface-visibility: hidden
arrives through an animation. Each rescan adds to them and the per-frame
backstop reads the same set, so an animated-only flip card is caught by both.
Look for WebGL canvases in the whole document: a render's shader-transition
compositor appends its canvas to body, while preview puts it inside the root.
Turn page blur off on routes that cannot host it. The HDR and layered
compositors screenshot each DOM layer after hiding the others, which does
not reach blur's canvases, and a distributed (cloud or Lambda) chunk render
has no Studio preview to match and no guarantee of html-in-canvas in the
worker's Chrome. The producer sets pageMotionBlur to hdr-render,
layered-composite or distributed-render there, the distributed plan's
compile-time hint no longer forces screenshot capture, and the runtime says
which route turned it off.
Find the same movers whatever time a scan starts: the baseline is read at
t=0, so each worker of a parallel render picks the same set. An ancestor's
pan is not counted as its child's move.
Folded review fixes: R5, R10, R15, R19.
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
html-in-canvas captures a few things wrongly, and a wrong blurred frame is
worse than a sharp one. Add a defect list the runtime checks before it wraps
a composition: a WebGL or WebGPU canvas (captured frozen), backface-hidden
faces in a 3D context (drawn mirrored), an iframe (reloaded by the wrap and
not recoverable), a transform or opacity below 1 on the root (ignored by the
capture) and a canvas that is already a capture canvas (nesting hangs the
renderer). Any hit turns blur off for the whole composition and the runtime
says which one.
The backface scan reads rules nested inside a style rule (CSS nesting),
sheets pulled in with @import and adopted style sheets. Shadow DOM stays a
documented limit.
A face whose backface-visibility: hidden arrives through an animation counts
too: the scan also takes the elements of running CSS or WAAPI animations that
key it, and elements naming a @keyframes rule that does, as candidates.
Support Chrome 157's html-in-canvas API alongside the older one: a canvas
marked content="drawable" with a drawable attribute on captured children
replaces layoutsubtree. Blur, the VFX runtime and Studio's browser check
detect both through the new @hyperframes/core/html-in-canvas subpath.
Folded review fixes: R14, R9 (the iframe defect entry).
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
html-in-canvas captures a WebGL or WebGPU canvas frozen or black, so the
runtime turns blur off when the composition draws one. Two cases slipped
past the tracker that spots them: a canvas handed to
transferControlToOffscreen (its context is created out of reach) and a
context made by an author script that runs before the runtime.
- The tracker is a self-contained function that also wraps
transferControlToOffscreen (a transferred canvas counts as accelerated)
and keeps its registry on window so independent installs agree.
- The engine evaluates it on every new document, ahead of any page script.
- insertRuntimeTag (bundler and Studio preview) puts it at the very start of
<head> for pages that opt into blur; the runtime install is idempotent
with it.
- report() carries failedFrames, the cumulative count of frames that
rendered plain DOM because their capture failed or timed out.
failures keeps its streak meaning.
Core's unpacked size budget rises from 3903211 to 3930000 bytes: the runtime
lifecycle, the compositor and the tracker together pass the old ceiling.
Folded review fixes: R6, R17 (the report counter; the engine log line lands
with the runtime's reason codes).
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
Carry the runtime's blur through the render pipeline and the Studio render
panel.
- engine: installPageMotionBlur sets the page-side off switch and the render's
device scale before any author script runs; CaptureOptions.pageMotionBlur
lets a caller stand the page blur down.
- producer: a composition that asks for data-motion-blur on the root the
runtime would pick forces screenshot capture, because html-in-canvas needs
a real paint; the probe session carries the request's pageMotionBlur like
every other session.
- studio-server and Studio: the render panel sends previewMotionBlur, the
render route turns "off" into pageMotionBlur once, and the Studio dev server
and the CLI server both receive the final value.
- mover scan: a large element that is see-through while it moves (a sweep, a
light leak) is no mover, including a large image or canvas with alpha; the
scan skips elements moved by a 3D individual transform property.
- cli snapshot injects video frames before the seek settles, matching the
engine's order.
Folded review fixes: R1, R2, R3 (scan half), R7, R16, R18, T3 (render
adapters).
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
The per-workflow guides now route through /hyperframes, so the sidebar group
names what the reader wants to do rather than skills that no longer ship. The
guides no longer promise a confirmation round the retired workflow skills ran
(the interview and brief); they say what to put in the request and when to ask
for a proposal or storyboard first. URLs are unchanged.
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* refactor(skills): fold six domain skills into hyperframes and media-use guides
hyperframes-core, -animation, -keyframes, -creative and -studio move to
skills/hyperframes/{core,animation,keyframes,creative,studio}/, and
hyperframes-audio moves to skills/media-use/mix/. Each moved SKILL.md becomes
a frontmatter-free GUIDE.md so installers that scan recursively for SKILL.md
no longer see them as skills; the published list goes from 21 to 15.
Relative links, code paths and skill-name mentions are rewritten to the new
locations, the catalog surfaces, marketplace core-skills list, fallback core
set, module-copy targets and skills-manifest.json follow, and
`hyperframes skills update` now removes the retired skills (RETIRED_SKILLS)
when they are still installed.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(skills): slim the core entries and give media-use a HeyGen and a free path
The four entries every run opens are rewritten to what an agent cannot infer:
hyperframes 21.8 KB to 7.7 KB, hyperframes-cli 24.8 KB to 6.9 KB,
hyperframes-registry 9.9 KB to 4.3 KB, media-use 9.7 KB to 4.9 KB. Detail moved
whole into references (routing close calls and creator edits, CLI pin upgrades,
catalog search tiers, CLI conventions and feedback reports).
- hyperframes: state table, routing, the render contract, a terminal work loop
(check, snapshot, preview before render), the desktop-app hand-off, and a
table of the folded guides.
- media-use: one table of HeyGen versus free local tools per need; HeyGen
sign-in, credits and the CLI's video-agent, avatar, clone, lipsync, translate
and clipping commands; every free path works with no account.
- scripts/check-core-skills.mjs holds entries and descriptions to budget and
checks every taught `npx hyperframes` command and path; runs in lint.
- Core entries are exempt from the formatter: its table padding cost ~1.5 KB
per entry on every run.
- Content tests follow the text to its new home.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): plugin runs skip skills update; media-use explains HeyGen sign-in and TTS setup
The smoke eval showed a presenter request never reaching media-use and the
free voiceover failing on a missing kokoro-onnx; plugin runs also ran the
skills update the plugin rules forbid.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore(skills): drop a committed Python bytecode file and ignore bytecode
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(skills): align the entry with the Studio skills and keep the workflows' measured facts
The entry drops routing and the intent interview and carries the contract, the
editable-project rules and the work loop. Studio's format, runtimes, motion,
look and captions references come over for terminal use, frame presets move to
assets/, media-use gains Studio's mixing and local-command references, and
film.md and overlays.md keep what the workflow skills measured.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(skills)!: retire the workflow skills and folded guides; publish seven knowledge skills
The skill pack now publishes exactly seven skills - hyperframes,
hyperframes-cli, hyperframes-registry, media-use, figma, slideshow and
remotion-to-hyperframes - carrying knowledge, not process. The eight
workflow skills, the folded domain guides and the routing/interview
references are deleted; their measured facts already live in the
hyperframes references (format, runtimes, motion, look, captions, film,
overlays) and media-use's mixing reference.
- bake-basemap.mjs and locate.mjs move to skills/hyperframes/scripts/
(bake-basemap gains --help); extract-audio-data.py moves to
media-use/audio/scripts/ for the caption motion reference.
- skills update: the eight workflows join RETIRED_SKILLS so existing
installs prune them (by lock attribution only, since their names are
generic); naming a retired skill reports where its knowledge went and
succeeds instead of failing as unknown.
- Capture prompts point at the hyperframes skill's references/film.md;
the live-map warning names the new bake-basemap path.
- Content tests that pinned knowledge are retargeted to the kept
references; tests of deleted workflows, scripts and process go.
- Catalog surfaces, templates and docs describe the seven skills.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): resolve the local.md captions link and allow capture's output files in citations
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): regenerate the skills manifest from tracked files only
The committed media-use hash included an ignored __pycache__ bytecode file present
in the local checkout, so CI's clean checkout hashed differently.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(docs): restore generated catalog pages and docs navigation to main
Catalog PRs commit item sources only; CI regenerates the pages from the edited
registry sources. docs/docs.json is catalog navigation, so its Workflows group
rename is left for a non-catalog PR.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* build(cli): raise the CLI size budget for the skill references it bundles
The CLI bundles skills/hyperframes, which now carries the folded domain guides and
frame presets (155 KB to 1.08 MB in git). Budgets are the CI-measured size plus 5%:
packed 7526634, unpacked 22593213.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* test: keep repo-sweep guards meaningful after the workflow skills left
The last shell ffprobe caller and most root index.html examples lived in retired
skills. The shell detector is now pinned on a literal command line, and the
structure-rule sweep needs more than 5 roots (11 remain) rather than 30.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(skills): slim figma, slideshow and remotion to the Studio shape; restore the storyboard board
slideshow 33.5 KB to 4.9, figma 18.2 to 6.2, remotion-to-hyperframes 11.0 to
4.9, each keeping the facts only it holds (the figma motion helpers, verify gate,
storyboard decoding and usage beacons; present and hand-off; the Remotion lint,
SSIM check and API map). The board page format from the storyboard recipe comes
back as hyperframes/references/storyboard.md: the desktop app's Storyboard tab
reads boards/storyboard.html by its frame, title and time classes. The entry and
the init template now say to snapshot the moments an edit changed before
replying, which the terminal eval found skipped in 22 of 38 runs.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): manifest without local bytecode; Studio's edit prompt names kept files
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): render when the person asks for the file or for no stops, not only after approval
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): keep the voiceover carve in media-use; presets stage their own fonts
Review follow-ups on #5651:
- carve.mjs and its tests move to skills/media-use/audio/scripts; the audio
docs, mixing.md and two Studio comments point there.
- code-editorial's FRAME.md says to copy its bundled WOFF2s and add the
@font-face blocks, instead of promising the retired build-frame.mjs.
- media-use states the host rule before the HeyGen sign-in steps.
- storyboard-format marks transition_in as a note the agent builds itself.
- Two stale citations (creator-editing-recipes.md, sub-compositions.md).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore(skills): regenerate the manifest without local bytecode
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): the CLI render step follows step 4's rule, not approval alone
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(ci): run the core-skills check from lint again after the main merge
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(skills): films that tell a story start from three storylines; maps use real imagery
Opus 5.5 arm D read motion.md and applied it, but its launch film was a
feature list (pain, three feature cards, proof, logo): film.md left story
to the model. And the map case never opened film.md, so it drew Natural
Earth outlines instead of baking satellite tiles.
- references/story.md: the generic half of Studio's moment-film story and
craft (one thing, three storylines with world and device, the hook,
beats, worlds, copy, brand, tells of a generated film). The person picks
a storyline, or the agent takes its recommendation when asked for the
file or for no stops.
- The entry's work loop names the storyline step for films over ~15 s; the
references table routes stories to story.md and maps to film.md.
- film.md points at story.md and gives maps and real places a section.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): story.md follows the entry: no voiceover unless asked
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): bake-basemap starts from a globe and stops waiting out its idle timeout
The Opus 5.5 map run spent 13 of its 38 minutes on parallel bakes that
stalled and 15 more recording 300 frames, after rewriting the script because
it could only start on a flat map.
- PROJECTION=globe with a seeded starfield and atmosphere, and FROM/PAN to
turn the globe to CENTER before the dive.
- Each frame polls for loaded tiles instead of waiting for an "idle" event
that had often already fired: 30 frames went from 198-219 s to 10-13 s
locally with byte-identical frames.
- Progress prints seconds a frame and minutes left; film.md shows the globe
call and says to run one bake at a time.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): render in the foreground; wait on a detached render's log, never pgrep -f
Both Opus 5.5 explainer runs started a detached render, then waited with
`until ! pgrep -f "hyperframes render"`, which matches its own loop: each
render took 88-117 s and the wait sat for the tool's full 10 minutes.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(skills): route each kind of film; storylines carry look and sound; add review and sound
The entry routes by request (launch, explainer, music, data, vertical, 3D,
character) to a self-contained file under references/types/, and a film
over ~15 s now gets music and sounds on its visible actions unless the
person asked for silence. Each storyline names its look and its sound, so
picking one picks both; STORYBOARD.md frames carry `reads` and `sound`.
review.md is the whole-film pass: the check warnings, the review sheets,
seven questions to look for, P0/P1/P2 notes and REVIEW.md. motion.md gains
eases by role, six title entrances, a reading-time rule and nine measured
transitions; look.md a design plan before the first scene, with presets
kept for decks and named picks. story.md adds the second default that
generated design falls into, with what to do instead.
The blank template and format.md no longer teach the fade-and-slide-up
entrance on a dark canvas.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): check warns on unread lines, dead stretches, small type, empty frame 0 and silence
A coding agent cannot watch its film, so check now measures what a viewer
notices over time. A new `craft` section holds warnings only (exit 0
unless --strict), sampled on the dense 8 fps grid the overlap pass
already seeks, skipped for --range checks:
- line_unread: a reading-size line (>= 30 px tall at 1080) on screen
under 0.4 s for 1-2 words, else under 3/4 of 0.5 s + 0.3 s a word;
the film's last line needs 2 s. Short lines are re-measured between
samples to ~0.01 s before judging. Captions-track and
data-layout-ignore text are exempt.
- dead_stretch: more than 3 s where the keepsMoving liveness signature
does not change, outside the final 3 s and any data-hold element;
quiet when sweep_static fires or a whole-composition keepsMoving
assertion owns the limit.
- text_below_floor: text under 2.6% of the short side (28 px on 1080)
held 0.5 s, one finding per file listing every element.
- first_frame_empty: frame 0 paints only the background.
- film_silent: over 15 s with no unmuted audio or audible video.
Each message names the fix and the skill reference to read. keepsMoving
now also leaves the film's last 3 s (a still end card) out of its
static runs, so the two agree.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): snapshot --review captures a beat sheet, seam strips and a phone sheet
snapshot --review picks the frames that show a film over time instead of
evenly spaced ones, in the same single browser session:
- review-beats.jpg: one settled frame per scene at its start + 0.3 s,
plus one per timed data-beat="name" element, labelled time and id.
Scenes are the composition's top-level timed clips and sub-composition
hosts, read like `timeline` reads rows (captions and audio excluded)
and timed by the runtime's start resolver over the live page.
- review-seam-NN.jpg: frames at -0.2, -0.1, 0, +0.1, +0.2 s around
every cut (each scene start, and any scene end no scene starts at).
- review-phone.jpg: the beat frames at 360 px wide.
--composition scopes the review to one sub-composition; with fewer than
two scenes the beats spread evenly and a note says so. snapshot also
gains --json: one envelope with every frame's path and time and, under
--review, the scenes, beats and seams behind the sheets; progress moves
to stderr.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): craft warnings stay out of --strict; a hold gets a real next action, not a caret
Craft findings (unread lines, dead stretches, small type, an empty first
frame, silence) are judgment calls a film may break on purpose, so
--strict keeps gating defects only. The check reference no longer
suggests a blinking caret to keep a still frame moving.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): review spreads beats when every scene starts together; map credits skip the type floor
A map film whose basemap and overlay both start at 0 s got one beat
frame; the even spread now keys on distinct scene starts. film.md tells
agents to mark the imagery credit data-layout-ignore and back it, since
the new type floor flagged it on the first real run.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): review frames each scene at its middle too, not only 0.3 s in
On the first real explainer, 0.3 s into a scene showed a bare colour
field before its title entered. Each scene longer than about a second
and a half now adds a frame at its middle, so the sheet shows what the
scene builds.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(skills): a film is built from a numeric SPEC.md, with defaults banned by name and a look test
Studied in full: flxrnc's Motion Lab (vague vs directed prompt), 0xMovez's
course, EianLu's riso skill, agentchengfeng's style samples and the two
Opus 5.5 guides. Their common point: adjectives are where the default look
lives, so the agent now writes SPEC.md before any scene (GOAL, OUTPUT,
TIME, LOOK, MOTION, SHOTS, BANNED, QA), with a number for every adjective.
spec.md bans the defaults by name, unless the brand or reference uses
them (a dark void with small things floating, glow and particles, the
cream-serif-mono-label look, fade-and-slide-up, corner brackets), and
grows that list after the first render. Each storyline gets a 3 s test of
its hero shot before the pick; a look is written as the rules of one
medium. Composition numbers: the hero fills 40% of the frame, one hero
move per shot, the shot's information in within 2.5 s then a held poster.
motion.md adds springs by damping ratio, shorter exits, overlap, stagger
and the beat grid; review.md reviews as a harsh director and grows BANNED;
vertical.md takes the riso skill's measured safe band; explainer.md the
shapes of information.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): story before look; every object reads as itself; posters only for info cards
The first spec-driven C02 swapped the dark default for a flat Swiss
poster, and the story went with it: two black bars nobody read as mirrors,
slogans instead of cause and effect, a storyline picked for its picture.
The spec now serves the story: a storyline is picked for the story a
stranger would follow, the look test asks first whether a stranger
understands what they see, and SPEC.md gains an OBJECTS line giving each
thing the material and details that make it read without a label. The
2.5 s held-poster rule is limited to information cards; in a story film
every shot causes the next. Explainers start from a person or situation
the viewer knows. Review names every object as a stranger would and
treats one that needs explaining as a P0.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): storylines are three films in words, each with a device and a form
The three storylines are pitched as text and the person picks from them;
nothing is built for a storyline before the pick, so the 3 s hero-shot
look test is gone. Ideas pair a device with a form, five or more first,
keeping the three that differ most, and the one stop states the rest in
an assumption line that names every paid item beside its free option.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* ci: list main's new embedded-captions hf-cli test among the retired workflow files
The merge with main removed it with the rest of the retired skill, and the
deletion guard rejects any removal it has not been told about.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(core): keep the audio preset comment within core's size budget
The comment's new citation to media-use's mixing reference made the published
package 14 bytes larger than its budget, which sits at main's exact size.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): craft keeps its own 8 fps grid; comments back within the checks
Review on #5651 found that the merge moved craft onto the overlap pass's
times, which main caps at 180, so a brief line in a long film fell between
samples and line_unread missed it. Craft now builds its own grid at 8 fps up to
600 samples, as before the merge, and a 60 s film with a line shown for 0.2 s
pins it.
The Comments check passes again: REVIEW.md and STORYLINES.md join the files a
skill writes into a project, the comment lines this branch added come back
under each file's share at the merge-base, and the audio preset comment drops
a word the check reads as history.
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(engine): a muted video no longer plays at full volume in renders that skip the browser probe
A <video data-has-audio="true" muted> with an audio stream is silent in
preview and in renders that run the browser probe, but a render that
skipped the probe mixed it at volume 1.0, ignoring data-volume. The audio
parser drops muted videos; the video parser kept hasAudio from
data-has-audio alone, so the auto-detect fallback re-added the sound.
The video parser now asks the same isAudibleVideoElement check.
* style(producer): format the muted-video parity test
Make the lifecycle from the previous change safe for real compositions.
- Frames are drawn in device pixels: the canvases and the compositor buffers
follow the page's scale, so a 2x render is not blurred at half resolution.
- A render never gives up mid-video: a frame whose capture throws is plain
DOM and the next one wraps again; in Studio, three capture exceptions in a
row turn blur off. (Paint timeouts render plain and never count; they are
handled in the previous change.)
- A hidden preview tab rests on plain DOM instead of waiting on paints that
never come.
- Timeline callbacks no longer fire during the mover scan, even with
data-motion-blur-events="on".
- Wrapping and unwrapping the root no longer reload iframes, restart CSS
animations or leak timing invalidations: each move runs through the
runtime's record filter and the adapters re-seek right after it. The waapi
adapter drops animations Chrome cancelled, limited to browser-created CSS
animations (a small, separate hunk in adapters/waapi.ts).
- Studio preview holds the playback seek while a blurred frame is in flight,
so the poses and both captures belong to the same time.
- The host exposes now() for the re-seek.
- Tests for the wrap, two-phase and painting paths, and for the guards that
survived a deliberate break.
Folded review fixes: R4, R8, R9 (runtime, init and waapi parts), R11, T1.
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
* feat(core): capture motion blur pose, discovery and compositor
Add the three pure modules capture motion blur is built from: pose (an
element's box as an affine in composition space, and the poses between two
times), discover (find the elements that move fast enough for a smear to
show) and composite (redraw each mover along its path into one frame).
Nothing calls them yet; the runtime that wires them up lands in the next
change.
The pose reads the individual translate, rotate and scale properties in CSS
order, adds the offsetParent border back and subtracts ancestor scroll, so it
matches getBoundingClientRect for a box centered with translate: -50% -50%.
Folded review fixes: R3 (pose half; the scan's 3D skip for individual
transforms lands with the mover scan).
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* feat(core): capture motion blur runtime lifecycle
Wire the pure modules into the composition runtime. A root that opts in with
data-motion-blur wraps itself in a capture canvas and, per frame, redraws
its fast movers along their paths into an output canvas. The runtime turns
itself off, and says why, when the browser lacks html-in-canvas, when a
WebGL or nested capture canvas is present, or when the engine's own
sub-frame blur is on. It unwraps when paused so Studio can select inside.
A frame whose paint never comes within 2000 ms is rendered as plain DOM
(root back in the page, movers revealed), counted in report().failedFrames
and never counted toward giving up.
The page-side off switch, PAGE_MOTION_BLUR_OFF_GLOBAL, is exported from
@hyperframes/core for the engine to set.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(lint): check no longer fails projects that inline the GSAP library
non_deterministic_code skips a script that is GSAP itself (producer inlining comment
or GSAP banner), using the detector missing_gsap_script already had, now one shared
helper. The author's own Math.random/Date.now calls are still flagged.
* fix(lint): only GSAP's own code is exempt from the determinism check
Exempt the parsed GSAP UMD statement that carries its banner, not every script that
mentions GreenSock. Strings, comments, spoofed banners and author code before or after
the library in the same script are checked again.
* refactor(lint): name the GSAP blanking step instead of commenting it
* fix(lint): exempt only GSAP's verified wrapper call from the determinism check
Recognise GSAP's UMD wrapper by structure (factory IIFE called with this and a function,
with a real define.amd member) anywhere in the tree, and blank only that call. Author code
joined by a comma or wrapped in an outer IIFE is checked; define.amd text is not enough.
* fix(lint): a bundle wrapping GSAP no longer inherits its determinism exemption
The define.amd check reads only the wrapper's own body, and each GSAP banner belongs to the
wrapper right after it, else the innermost one containing it. An outer wrapper or UMD bundle
around pasted GSAP keeps its author code checked.
* fix(lint): a banner inside a bundle no longer exempts the bundle around GSAP
A banner introduces the wrapper after it across whitespace, ! ; ( and comments, so producer
comments and statement separators bind to GSAP itself. The fallback owner for a banner
inside a wrapper must wrap no other wrapper, so a bundle around GSAP never inherits it.
* fix(lint): author code bundled with GSAP keeps its randomness checked
A GSAP banner now exempts only the UMD wrapper call that starts right after
it (skipping whitespace, comments, `!`, `;` and `(`). The fallback that gave
a banner to the innermost enclosing wrapper is gone: Rollup's UMD output for
an ESM import of GSAP keeps the banner inside the one factory it shares with
author code, so that fallback exempted the whole bundle, author
Math.random() and Date.now() included.
Tradeoff: unminified dist/gsap.js pasted inline has its banner inside its
own factory, so it is flagged again, as on main. The minified gsap.min.js
and producer-inlined copies stay exempt.
* test(lint): the Rollup bundle case reuses the GSAP banner constant
* fix(lint): a script name that mentions gsap no longer exempts the bundle it heads
The producer's inlined-source comment now marks library code only when its
URL path is an npm gsap dist file (gsap or gsap@<version>, then /dist/<file>.js,
query and hash ignored). A Rollup/Terser bundle of GSAP plus author code
inlined from gsap-composition.js, scene.js?vendor=gsap or another domain's
gsap-demo/scene.js keeps its own Math.random() and Date.now() checked.
The AAC true-peak correction assumed each dB of attenuation lowers the
encoded peak by a dB, aiming 0.5 dB under the -1 dBTP ceiling every pass.
FFmpeg 8.1's AAC encoder can hold the peak flat over a few dB of
attenuation (a production render measured -0.5, -0.6, -0.6 dBFS at -3.4,
-4.4, -5.3 dB), so three passes ran out and the render failed. Each pass
now aims twice as far below the ceiling (0.5, 1, 2 dB). The first pass is
unchanged; exhausting three passes still fails the render.
* perf(producer): cloud render chunks start their capture browsers while video frames extract
* refactor(producer): say the chunk extraction overlap in names, not comments
* feat(cli): cloud render uploads heavy media as HeyGen assets
Media of 5 MB or more leaves the project zip: it uploads through the
direct-upload flow as a HeyGen asset and the zip's HTML and CSS point at
the asset URL. Uploads are cached by content hash and re-checked before
reuse. Media the asset API refuses or caps is re-encoded at visually
lossless quality first. The zip must end under 200 MB or the command
stops and names the largest files left.
* feat(cli): cloud render offloads GIFs, still images and script-referenced media
Animated GIFs in the entry composition go through the renderer's own GIF
preparation and upload as the resulting video. Still AVIF, GIF, BMP and
TIFF images convert to PNG. Plain string literals in scripts and JSON that
name offloaded media point at the asset URL. ffprobe calls end their
options with "--", and two comment lines are trimmed for the ratchet.
* feat(cli): cloud render uploads only the parts of long media a composition plays
A video or audio file that has to be re-encoded is cut to the union of the
windows its clips play, with half a second of margin on the source's frame
grid, and the staged composition's media starts shift to match. Clips in
sub-compositions, looping clips, speed ramps and files a script names keep
the full file.
* fix(cli): cloud render no longer decodes media files as text when rewriting references
* fix(cli): cloud render keeps media chosen at render time and encodes the staged bytes
Only media a reference points at is offloaded, so a file picked through
--variables or built by a script stays in the zip. Variable-bound clips and
clips with a media fragment keep their full length. Entry GIFs resolve the way
the renderer rebases a nested entry and are prepared only from 5 MB, and every
encode reads the staged bytes its cache key hashes. Query strings and media
fragments survive the rewrite, a null cache file reads as empty, and the docs
say asset URLs are public and unsigned.
* fix(cli): cloud render comments cite only what this repo holds
* fix(cli): cloud render handles media over 2 GiB and explains a too-large upload
Media of 5 MB or more stays on disk until it is used: it is hashed and uploaded
as a stream, and a file edited after staging stops the run instead of caching
the wrong upload. Files left in the zip are sized before any is read, so a
4 GB file reports "too large" instead of crashing.
The asset API rejects an over-limit upload as invalid_parameter with "File too
large"; that now gets the dry-run hint instead of the --help one.
* fix(cli): cloud render checks and reads each staged file through one handle
Staging, the on-disk staged read and the re-encoded output read each opened the
path twice: once to stat, once to read. A file swapped or edited between the two
could be read without the size or change check seeing it. Each now opens once,
checks the open file's stat before and after the read, and reads from it.
* test(cli): a heavy file a script builds from split strings stays in the cloud render zip
* fix(cli): cloud render keeps media scripts may reach and never trims clips scripts retime
- An offloaded file keeps its zip copy when script code holds both its stem and
extension, so a path joined at runtime still loads.
- Entry GIFs are prepared from a private copy checked against the staged file,
so the GIF cache never stores a later edit.
- No cut when any script can write a clip's timing attributes, dataset,
currentTime, playbackRate, src or loop.
* perf(producer): multi-worker mp4 renders start encoding while frames are still being captured
On the multi-worker disk capture path the encoder used to start only after
every worker had finished. It now starts with capture and is fed frame 0, 1,
2... in order as each frame file lands, with the same encoder arguments, so
the output is byte-identical and the encode left after capture is shorter.
Off for drawElement capture (a failed self-check discards every frame) and
for png/gif/webm/mov/HDR/GPU/chunked/HLS output. If the overlapped encode
fails, the frames are encoded after capture exactly as before.
* perf(producer): encode during capture only when a core is spare for the encoder
With every core busy capturing, the encoder only slowed capture down, so the
overlap now runs only when the render uses fewer workers than the CPU count
worker sizing sees. Also pins the reviewed producer test-lane source hash and
drops an added comment line.
* refactor(producer): shorter comments on the encode-while-capturing helpers
* fix(studio-server): keep transparency in proxies of VP8/VP9 alpha webm
A webm whose alpha lives in the VP8/VP9 side channel (alpha_mode=1,
pix_fmt yuv420p) was proxied to opaque H.264, and the VP8 proxy path
decoded it with ffmpeg's native decoder, which drops alpha. The alpha
rule and the alpha-reading decoder now live once in
@hyperframes/parsers/media-alpha, used by engine and studio-server.
* build(parsers): raise the packed size budget for the media-alpha subpath
parsers already sat at its packed budget on main; the shared alpha rule
adds 676 B of JS and 803 B of declarations (190958 B packed in CI).
* perf(core): a project with many composition variables builds without a style scan per variable
* perf(core): a build checks each project path once, however often a scene is mounted
* perf(core): collecting a page's characters for fonts no longer copies the page into arrays
* perf(core): a warm font cache builds the page without asking Google Fonts again
The woff2 files were cached on disk, but the CSS lookup that names them was
only cached in memory. Each new process asked fonts.googleapis.com first, and
offline the lookup failed before the cached woff2 files were ever reached.
The lookup is now kept on disk beside them, keyed by its request URL. Only a
lookup that named faces is kept, written aside and renamed into place.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* perf(core): a build resolves and reads each inlined asset once, however often a scene is mounted
Inlining a relative asset as a data: URL ran isSafePath on it, which realpaths
the path and every ancestor, then statted, read and base64-encoded the file.
It did so for every url() of every mount, so a scene mounted 160 times paid
for the same few files 160 times. The answers are now kept for the build: a
path's safety is shared with the entry-path checks, and an asset's data: URL
(or its too-large warning) is kept by file path.
The bundler also tells onRead about each path once per build. The preview's
read tracker realpaths and readlinks every path it is told about, and it was
told again for every mount.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* perf(core): a scene mounted many times is read, parsed and scoped once per build
Every mount of a sub-composition re-read its file, re-stamped its hf ids,
parsed it four times (twice to check it is usable, twice to mount it) and
re-ran postcss over each of its stylesheets, though only the scope selector,
ids and variables differ between mounts.
- The bundler reads and stamps each scene file once per build.
- inlineSubCompositions parses each file once, answers the usability check
from that parse, and gives each mount its own clone to edit.
- createCompositionCssScoper scopes a sheet once around a placeholder scope
shaped like the real ones, then substitutes each mount's scope. It falls
back to scopeCssToComposition for any scope or id outside [\w.:-], a scope
equal to the authored id, a sheet holding the placeholder, or a selector
kept raw for a comment.
A golden test pins the preview and render pages of a representative project
(goldens written by the build before this change), and a fuzz test compares
the scoper with scopeCssToComposition over random sheets and scopes.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Miguel Angel Simon Sierra <miguel.sierra@heygen.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* perf(core): a project with many composition variables builds without a style scan per variable
* perf(core): a build checks each project path once, however often a scene is mounted
* perf(core): collecting a page's characters for fonts no longer copies the page into arrays
* test(cli): drop the orphan cleanup tests that read the runner's live process table
* fix(cli): a preview refused inside a host app no longer kills orphaned Chrome
* ci: a hung Windows test run keeps its log, and commit hooks lint and format .mjs/.cjs files
* ci: the Windows test job leaves room for setup before its step limits
Add the three pure modules capture motion blur is built from: pose (an
element's box as an affine in composition space, and the poses between two
times), discover (find the elements that move fast enough for a smear to
show) and composite (redraw each mover along its path into one frame).
Nothing calls them yet; the runtime that wires them up lands in the next
change.
The pose reads the individual translate, rotate and scale properties in CSS
order, adds the offsetParent border back and subtracts ancestor scroll, so it
matches getBoundingClientRect for a box centered with translate: -50% -50%.
Folded review fixes: R3 (pose half; the scan's 3D skip for individual
transforms lands with the mover scan).
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(core): grade an image that appears after its scene starts hidden
A graded <img> was only picked up if it was visible when grading scanned the page. An image in a
layer revealed later by a timeline, or in a sub-composition whose host starts later, stayed
ungraded in snapshot and render. Each draw pass now adopts graded media that has come into view.
* fix(core): keep the grading file's comment share flat
* fix(core): keep a grade switched off live from coming back on the next frame
* fix(cli): a disk that fills mid-unzip fails the browser install at once instead of hanging
* test(cli): the unzip tests hold on Windows, and a cancel stops unzip
* fix(inspector): a component report shows the compositions and hosts that time a matched element
* fix(inspector): resolve mounted roots and Windows ancestry
* fix(inspector): label repeated mounts as source candidates
* fix(inspector): budget candidate context after primary snippets
* fix(inspector): a component report names a snippet's file and lines only in its id
* fix(inspector): drop the comment the ratchet refuses
* fix(inspector): prioritize unqueried co-targets in compact reports
* chore: release v0.8.182
* chore: list #5953 in the v0.8.182 notes
* chore: add #5843, #5932 and #5945 to the v0.8.182 notes
* chore: add #5923, #5943, #5956, #5960, #5961, #5963, #5964 and #5966 to the v0.8.182 notes
* chore: add #5972, #5977, #5981 and #5982 to the v0.8.182 notes
* chore: add the 13 PRs merged since the last refresh to the v0.8.182 notes
* chore: add #6003, #6005 and #5579 to the v0.8.182 notes
* chore: add #6006 to the v0.8.182 notes
* fix(cli): a first render on a nearly full disk refuses at once instead of hanging
* fix(cli): name the low-disk guards instead of commenting them
* fix(cli): a killed install's leftovers never count against the free-space check
* test(cli): the render command's mocks carry the disk check it now runs first
* fix(lint): place each scene's audio at its host's start in the audio overlap check
* fix(lint): drop internal tags from two comments
* fix(lint): check audio in scenes chained by a reference start
structure.catalog.test.ts found root compositions with
endsWith("/index.html"), which never matches the backslash paths
path.join returns on Windows, so it counted 0 roots and failed.
It now compares basename(file) to "index.html".
The stripJsStringLiterals scaling test read one scan's CPU time from
process.cpuUsage, which Windows updates in ~15.6ms ticks. The 10k-input
scan finished inside one tick, read 0ms and made the ratio Infinity.
Each sample now repeats the scan until it has spent 50ms and divides
by the run count; the asserted ratio is unchanged.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(inspector): a component report under budget keeps the animation code before CSS candidates
Snippets filled maxBytes in file order, so CSS rules near the top of a file used the budget and the animation code further down was omitted. They now fill it by kind: animation creation, caller, timeline attachment, HTML opening tag, source dependency, then matched CSS candidates; the kept snippets are still returned in file order.
* fix(inspector): preserve explicit snippet request order
* fix(engine): cloud renders fail instead of going silent when an audio read is cut off
* fix(engine): cloud audio reads of a file whose audio ends early no longer fail
* test(engine): the audio link test's media setup passes the ffmpeg argv contract
* perf(producer): cloud render chunks hand Chrome each video near the size it is shown
A chunk now measures the largest box each video is drawn in over its own frames,
in device pixels, before extracting that video's frames, and halves the frames
while they still cover that box. Graded videos, pages with a canvas, and
object-fit none/scale-down keep full-size frames.
* docs(engine): shorten the shown-box measurement comment
* refactor(engine): name the halving filter instead of commenting it
* test(producer): end ffprobe options before the frame path in the shown-size tests
* feat(aws-lambda): a cloud render mixes its audio beside its chunks
* fix(aws-lambda): the audio task keeps a worker free and runs only when the plan left audio out
* test(aws-lambda): the audio task's free worker holds for a reserved pool of 1 and 2
* refactor(aws-lambda): one helper points every handler at the bundled Chrome
* fix(catalog): token-themed component previews no longer render black
Seven components read theme tokens with no fallback, so with no theme
pack they painted dark text on a black canvas, both when installed and
in their live Catalog preview. They now carry exact-value fallbacks from
their own demos, as the theme contract requires, and a test fails any
registry item that reads a contract token bare.
The Catalog's snippet preview wrapper also copies the tokens each demo
sets on its mount, and paints the demo's background, so the preview
wears the demo's theme.
* fix(catalog): trim the preview theme helper comment
* test(catalog): strip HTML comments with the parser in the theme fallback gate
* feat(producer): a cloud plan can leave its audio to be mixed beside its chunks
* fix(producer): separate plan audio keeps the sound of videos whose src a script sets
* refactor(producer): shorter plan comments on deferred audio and extraction
* fix(cli): a finished capture exits even when a site's DNS never answers
Capture resolved every page request with the system lookup and gave up after
10 s, but a system lookup cannot be cancelled: the abandoned one kept the
process alive until the OS resolver timed out, so a capture that had printed
its result could run past the app's 5-minute cap and lose its folder.
Resolve through c-ares instead, as hostAnswers does, and cancel it at the
same 10 s deadline.
* test(cli): a capture lookup that answers must exit right after too
Dropping the timer cleanup kept every successful lookup alive 10 s and no test noticed.
* fix(cli): capture falls back to the system lookup and exits once its result is out
c-ares answers first. When it has no answer, the system lookup (hosts file, OS-only DNS)
runs in a child process killed at the 10 s deadline, cached per host, because Node waits
for an in-process system lookup even in process.exit. Capture then asks the CLI to exit
100 ms after printing its result, as render does.
* fix(cli): capture reads the hosts file first and kills its lookup child on exit
The hosts file answers before DNS, as the system lookup does (overrides and 0.0.0.0
blocks keep working). A host whose system answer is cached skips c-ares, a caller that
joins a running lookup keeps its own deadline, and a pending lookup child is killed when
the CLI exits.
* fix(cli): capture's forced exit waits until a slow reader has the whole result
The root exit now waits for stdout and stderr to hand over what is already
written before it calls process.exit, so a large JSON result on a slow pipe
arrives complete. A hostname listed several times in the hosts file now has
every address checked, not only the first.
* fix(cli): keep process exit in the CLI root and let test fixtures exit
The output wait is now a helper that only waits; cli.ts still owns the
process.exit. The process-ownership check treats test fixtures like tests,
since each one runs as its own process.
* chore(cli): run the process-ownership check before every commit
A process.exit outside a process root used to pass the local hooks and fail
only in CI's Preflight.
Xuanru Li