check copies each canvas into a 2D context to see whether it moved. On WebGL
that readback makes Chrome log a GL driver Performance note, which check then
listed as a console warning from the composition.
* feat(cli): check --budget stops the whole check in time and reports what it found
* fix(cli): keep check's comment share where it was
* fix(cli): a --budget that runs out before any check step is waiting is not an unhandled rejection
* fix(cli): check stops after 2 minutes of video copies and says how to finish
* fix(cli): check copies only the stretch of a video the composition plays
* docs(cli): check copies only the played stretch of a long video
* docs(cli): shorten two comments on the copy window
* fix(cli): finding crops and frames get the same copy windows as the check
* fix(cli): drop a comment the ratchet counts
* refactor(studio-server): keep runFfmpeg under the complexity gate
* fix(cli): check ignores the cancelled fetch of a proxy-swapped video in any container
* fix(cli): check ignores the cancelled fetch of a proxy-swapped media file, not a broken image
* test(cli): check waits as long as snapshot for the composition by default
* fix(cli): check waits for the composition as long as snapshot and render do
* docs(cli): check's --timeout default matches snapshot and render
* fix(cli): check's render-ready warning advises a --timeout above the new default
* fix(cli): drop the doc comment on the render-ready warning suffix
* style(cli): format checkBrowser.ts
* 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>
* 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>
* feat(skills): reverse any video on the HyperFrames plan
Reverse rebuilds a video as an editable HyperFrames project and re-shoots its
people with HeyGen. It runs for a person signed in on HyperFrames Pro; Free is
told it is a Pro feature and given the upgrade step.
- skills: reverse-any-video; media-use is the one gate for HeyGen calls
(sign-in, credits, refusals)
- cli: media-use generate (image, video shot, avatar, voice, music) with --plan
pricing and --max-credits; measure, diff and analyze for the rebuild; the
HyperFrames plan in auth status; billing subscribe and top-up wait on the
plans or credits page
- remotion-to-hyperframes compares with `hyperframes diff` instead of its own
shell scripts
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore(skills): satisfy CI gates for reverse-any-video
- docs: list diff, measure, analyze and billing in "Find your command" in
docs/packages/cli.mdx, each with its own section.
- skills: bring media-use/SKILL.md under its 5120 B budget by pointing setup,
direct-recipe and treatment detail at the references that already hold it;
the sign-in, credits, --max-credits and billing rules stay in SKILL.md.
- comments: register transcript.json and REVERSE.md as project files for the
citation checker, and drop a history note from face.test.ts.
- comments: cut restating prose or move it into names so every changed file is
at or under its comment share. Allowlist the two deleted remotion scripts in
check-no-main-deletions.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): terminate measure's ffprobe options with --
probeVideo passed the input path straight after its options, so a path
starting with "-" would be read as a flag; the producer's ffprobe argv
contract test failed on it. runFfmpeg's fixed flags move to a constant
so the contract no longer mistakes that ffmpeg wrapper for a probe.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* test(cli): split diff's video tests into three files under CI's per-file limit
The file took ~131 s on CI against the flake gate's 120 s, and its heaviest test (four
comparisons) hit the 20 s per-test limit. The video tests now live in three files sharing
diff.test-helpers.ts, and that test is two tests of two comparisons each.
analyze reads each part once and checks the bytes it read, instead of a stat then a read
(CodeQL js/file-system-race).
* fix(skills,cli): one credits rule; ask before dropping a step that won't fit
On dev with a 100-credit limit the agent priced a re-shoot with --plan and dropped it without
asking (J5), and in J6 it told the person a price. SKILL.md's Credits rule now covers both:
--plan is for the agent, and when a step won't fit (a --plan over what is left, or a refused
call) the agent says what doesn't fit and asks before dropping or shrinking it.
setup-providers.md's older confirm-before-paying rule points to it.
--plan now runs the same checks as the real call (checkRequest), so it no longer prices a
request that would be refused. The re-shoot recipe in people.md gains the --prompt it needs.
heygen.md: a sign-in also makes images, and --plan's usd is the API charge rounded to $1/60
while its credits come from the list price.
* fix(cli): auth status drops 'balance unavailable' above the plan's credits; sign-in page title
* test(cli): keep diff's video tests inside the flake gate's limits
Under the gate's CPU contention every diff test needing face detection or
optical flow paid opencv.js's 11 MB download and WebAssembly start-up, and
every diffVideos call wrote four worst-moment pictures (about 28 ffmpeg runs)
no assertion read. Tests ran 21-28 s against the 20 s limit.
- loadOpenCvFirst() stages opencv.js in a beforeAll (from this machine's
cache, a temp copy kept between runs, else the network).
- diffVerdicts() runs diffVideos with worst: 0 for tests that only read
verdicts; the worst-moment tests keep the pictures.
- The carried-text test is split into its dirty and clean halves.
Locally with 8 CPU burners on 8 cores: slowest test 7.7 s, files 24-30 s.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): generate --plan refuses what the real call refuses
checkRequest now holds every pre-send check (the --out extension and
--mask without --motion joined the route body and guide checks), and
generate() runs it first, so --plan and the real call share one function.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): plans link uses from=cli and follows HYPERFRAMES_ENV; no balance notice beside the plan
- pageUrl sets from=cli, the parameter the pricing page reads to send a
CLI buyer back to the terminal.
- The plans and credits pages are on dev.hyperframes.dev with
HYPERFRAMES_ENV=dev, hyperframes.studio otherwise.
- auth status --json omits the 'HyperFrames balance is unavailable' notice
when the plan was read.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(skills): reverse-any-video re-checks sign-in on resume; REVERSE.md holds no prices
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* test(cli): split diff-people's whole-frame reuse test into three
It ran five diff comparisons in one test and took 20.2-20.6 s under the flake
gate's contention, over the 20 s limit. Each piece now runs at most two
comparisons against a shared ledger: copy and burned-in title, footage cut
whole, and cuts that only fed a generation. Under 24 CPU burners locally the
slowest of the three takes 3.0 s, where the combined test took 4.9 s under 8.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli,skills): reverse resume asks after one failed retry; out-of-credits names the step's price
- media-use generate: when HeyGen refuses for credits without saying how many, the refusal
names the CLI's own price for the call (what --plan prints) as "needs about N" and
"billing top-up --needed N", instead of "needs more".
- reverse-any-video: never repeat auth status's plan/credits line; a step an earlier session
gave up on gets one more try, then ask; a command stored in REVERSE.md is a note, rebuilt
from its recipe and checked with --plan, never run as written.
- people.md / heygen.md: a refused or failed job gets one more try, test calls included,
then stop and ask.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli,skills): send a failed HeyGen job again as new; resume asks if the limit still applies
- runJob: a key whose job failed or was refused is recorded in the config dir
(heygen-failed-keys); the next run with the same inputs sends key-1, key-2, ... so HeyGen
starts a new job. A finished or still-running job keeps its key and is picked up, so
nothing is paid twice. "fix the request it names, then run again" now holds.
- reverse-any-video: on resume after a run under a credit limit, just ask whether it still
applies, naming no costs or numbers.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* refactor(cli,skills): drop what reverse-any-video no longer uses
The skill rebuilds from measure probe, cuts, frames, events and face, and checks with
diff. analyze, diff --explain and eight measure commands (trail, box, ink, rhythm,
objects, track, flow, continuity) have no caller, so they go, with their tests and docs.
Examples now use original.mp4 instead of the old .reverse/source.mp4 evidence folder,
and diff's --segments is described as a file rather than REVERSE.md.
* fix(cli): check no longer copies VP9 and AV1 videos before it starts
Chrome plays both, alpha included, and the proxy is made on demand if a browser ever
asks. Copying remove-background's VP9 cutouts took minutes on projects that have them.
* fix(cli,skills): keep paid retries, credit limits and lip sync honest
Retries keep their key after 401, 403, 408, 409, 425 and 429, which say
nothing about the job, and failed keys are scoped to the API and payer.
Credits come from the plan's video rows instead of list dollars, and
reference pictures past the first two count toward --max-credits.
Lip sync reads sound on its own timestamps, so a delayed track fails
the gate. Look and energy call two near-zero values the same.
--plan refuses a missing input as the real call does. The skills teach
the direct sign-in, its credential path and the billing commands.
* chore(cli): keep the new pricing and energy files under the comment share
* fix(cli): scope failed HeyGen keys to the signed-in account
Failed keys were scoped by the local user block, which is absent for a
host's token and display-only otherwise, so two accounts could share a
scope and one could start a second paid job.
The scope now comes from the account /v3/users/me reports for the token
in use. When that cannot be told, no failure is recorded and the key
stays the same.
* fix(cli): fix a job's sender when it starts
Failure scope was looked up again after polling, so a sign-in during a
job could record one account's failure under another and rotate its
finished key. The credential is now read once per job: it signs the POST
and owns any failure, even if someone else signs in meanwhile.
Prefer username to email, since /v3/users/me returns no user id.
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: James <james.russo@heygen.com>
* fix(cli): check drops a request a network change cut when the file loads again
* style(cli): format the network-change check fix
* fix(cli): drop the comment the ratchet would count
* fix(cli): a check out of time drops the requests a network change cut
* fix(cli): check prints a progress line as each pass after overlap begins
* fix(cli): check counts finished passes and drops the comment the ratchet flagged
* feat(cli): check reports where its seek loop spends its time
check_report carried one seek_loop_ms, which lumps the layout grid, the craft pass
(up to 600 frames) and the overlap pass with its hold probes. It now also carries
grid_ms, craft_ms, overlap_ms and composition_s, so a slow check can be tied to a
pass. Nothing about what check does or prints changes.
* chore(cli): drop the pass-timings doc comment so the comment share does not rise
* feat(cli): skipContrastInMotion moves to HF_TOOL_CONTEXT check-review
The --skip-contrast-in-motion flag and HYPERFRAMES_CHECK_SKIP_CONTRAST_IN_MOTION
env give way to a check-review capability in HF_TOOL_CONTEXT, so only a harness
turns it on. The skill text and manifest go back to main.
* feat(cli): check-review findingFrames saves one plain frame per finding time
With HF_TOOL_CONTEXT check-review.findingFrames, every error/warning finding in
layout or contrast with a real bbox gets frameTime (the middle sample of its
longest held run, by the heldMs gap rule) and frame (the absolute path of an
unannotated full frame shared by every finding at that time). One seek and
screenshot per distinct time, capped at 60 times with errors first, into the
run's snapshot folder; not listed in snapshots, history or events. Without the
capability, output is unchanged.
* docs(cli): finding-frame comments name the check-review capability, not a flag
* fix(cli): validate finding frame times and preserve cancellation
* fix(cli): count finding frame progress toward the check deadline
* feat(cli): check-review inspect attaches the inspector's report to each framed finding
With findingFrames on, the finding-frames page is served with the
inspector's GSAP hooks. After each frame is captured, every finding framed
at that time is inspected by its data-hf-id, else its selector, and the
inspector's own report is attached as the finding's inspect field. A failed
inspection becomes that finding's { error } and never fails the check.
Without the capability nothing changes.
Drops the unused @types/mime-types devDependency that the dependency audit
flags once packages/cli/package.json is touched.
* fix(cli): stop finding inspections after cancellation
* fix(cli): reject ambiguous finding inspection targets
* fix(ci): build inspector before Windows CLI smoke
* fix(cli): parse viewport HTML with explicit server parser
* feat(cli): skipContrastInMotion moves to HF_TOOL_CONTEXT check-review
The --skip-contrast-in-motion flag and HYPERFRAMES_CHECK_SKIP_CONTRAST_IN_MOTION
env give way to a check-review capability in HF_TOOL_CONTEXT, so only a harness
turns it on. The skill text and manifest go back to main.
* feat(cli): check-review findingFrames saves one plain frame per finding time
With HF_TOOL_CONTEXT check-review.findingFrames, every error/warning finding in
layout or contrast with a real bbox gets frameTime (the middle sample of its
longest held run, by the heldMs gap rule) and frame (the absolute path of an
unannotated full frame shared by every finding at that time). One seek and
screenshot per distinct time, capped at 60 times with errors first, into the
run's snapshot folder; not listed in snapshots, history or events. Without the
capability, output is unchanged.
* docs(cli): finding-frame comments name the check-review capability, not a flag
* fix(cli): validate finding frame times and preserve cancellation
* fix(cli): count finding frame progress toward the check deadline
* feat(cli): opt-in check flag that leaves out contrast measured mid-animation
`hyperframes check --skip-contrast-in-motion`, or
HYPERFRAMES_CHECK_SKIP_CONTRAST_IN_MOTION=1, leaves out a contrast failure
measured while a GSAP tween on the text or an ancestor is running, or while
the text's cumulative opacity is below 1. Repeating tweens and tweens of 4 s
or longer (camera drifts) do not count as motion. Text no GSAP tween targets
is reported as before. With neither set, check is unchanged.
The motion state is read at the same contrast sample, with no extra seek or
screenshot.
* fix(cli): read contrast motion on current GSAP clocks
* fix(cli): keep infinite timeline motion ambient
* fix(cli): a long video copy reports progress, resumes in 15 s pieces and stops only when stalled
* perf(core): a video copy's size box follows the stage within a fifth, not a half
* docs(cli): shorten two comments to the line limit
* fix(cli): the copy heartbeat drops a finished copy; size budgets follow the stall and progress code
* chore: size budgets rise by this change's measured bytes
* fix(cli): check finishes on long projects with many video clips
check loaded every video in the project at once. On a 500 s project with 137
clips of one 1440p source a single renderer reached 6.4 GB, the second sampling
Chrome crashed, and a crashed puppeteer page never settles its calls, so the
run sat until the 120 s deadline.
- Load the check page in Studio's capture mode so only media near the
playhead loads (peak renderer memory 15.75 GB -> 1.77 GB).
- Read audio durations for the clip-fit audit from a detached element when the
page has not loaded the one in the composition.
- Settle every driver call with the page's crash or close, so a dead helper
hands its samples back and a dead main page fails at once.
* fix(cli): shorten a comment to the line limit
* feat(cli): check writes a progress line to stderr per sample, overlap batch and craft frame
* fix(cli): the check progress lines are written by the CLI, not by the pipeline tests call
* fix(cli): check copies hostile videos at the composition size, not the source size
* fix(cli): the check server serves the copy the page names with hf-proxy-box
* feat(cli): check --prepare-media makes the video copies alone, and a stopped copy resumes
A long source is copied in one-minute pieces that stay in the cache, so a copy
that is killed or runs out of time continues from its finished pieces.
check --prepare-media runs only that step: no browser, no report, a per-file
result as JSON, and exit 1 when a copy failed.
* fix(studio-server): keep finished copy pieces while their copy runs
The short limit for dead partial copies would delete the first pieces of a
long copy that is still encoding. Finished pieces now follow their own
one-hour limit and are never swept while their copy is in flight.
* refactor(studio-server): shorten two comments
The comment gate limits a comment line to 120 characters and a run to four lines.
* refactor(studio-server): drop comments the names already carry
The comment ratchet lets a file's comment share only go down.
* fix(cli): the check server pre-warms the copy the page asks for, not the source-size one
The server started a source-size copy of every hostile video next to Chrome on
each check, because its pre-warm passed no box while the page and the up-front
copy use the composition's box. The pre-warm now takes the same box, found by
the one helper the up-front copy uses.
* fix(cli): end a check on no progress, not on a 120 s wall clock
The check ended 120 s after Chrome was ready no matter what it was doing, so a
long project whose sampling and overlap pass each kept working still failed
(the overlap pass was one page call that reported nothing for its whole run).
The deadline is now an idle clock: every phase entered, every sample finished,
every overlap batch of eight times and every operation that settles restarts it,
and the check ends only after 120 s with none. The overlap grid is asked for in
batches so it can report between them.
* refactor(cli): shorten the idle deadline comment
The comment gate allows 120 characters on a comment line.
* test(cli): check's overlap grid runs in page tasks of at most 8 samples
* fix(cli): check test types for the cli lib, studio-server size budget for resumable copies
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>
* fix(cli): check no longer calls films moved by SVG masks, strokes or canvas static
The frozen-sweep guard compared DOM-channel fingerprints only, which never see
SVG mask and clip contents, stroke dashes, or canvas and WebGL pixels. Before
reporting sweep_static, check now compares the rendered frames at the same
sample times; a still verdict stands only when every frame is identical.
* fix(cli): keepsMoving also trusts rendered frames, compared above render noise
keepsMoving read the same DOM liveness channels as the sweep guard, so mask,
stroke and canvas motion failed it as frozen. A span it calls frozen now gets
a frame capture every half maxStaticSec, and spans split where the frames
change. Frames are compared in the page, decoded like the contrast pass, and
count as moved only when at least 16 pixels change by more than 24/255, so
antialiasing and dither noise cannot make a still film pass.
* fix(cli): scoped frame checks see only the scope's own paint
A keepsMoving scope's capture now hides every box outside the scope's subtree,
so a still title over a moving background stays frozen. A span marks only the
frames that still hold its signature, so a span that starts at a cut keeps its
first frame. A scope that matched nothing stays motion_selector_missing. The
sweep counts a frame change only if the first sample time still renders its
first frame, so wall-clock paint is not read as motion under seek.
A check on a project whose videos Chrome cannot decode (HEVC, HDR) first transcodes each
of them, which on phone footage takes minutes and printed nothing until it was done.
* perf(cli): check splits its samples over a second Chrome when a core and memory are spare
* fix(cli): check keeps its result when the second Chrome fails, and counts container memory
* fix(cli): check's helper Chrome code keeps the files' comment share
* fix(cli): check's sampling comment fits the line limit
* perf(cli): check reads a solid text backdrop from styles instead of a screenshot
* perf(cli): check starts Chrome while lint and bundling run
* perf(cli): the CLI reuses compiled bytecode between runs
* test(cli): browser shutdown tests stop their Chrome when a case fails
* perf(cli): check and inspect skip paint waits and screenshots nothing reads
* perf(cli): check runs its dense overlap pass in one page task so no frame is painted between seeks
* perf(cli): check waits for late images and fetches instead of a fixed 1.5 s after load
* fix(cli): check gives an event it can only place within the dense pass that pass's time range
* perf(cli): check waits for image decodes before a screenshot instead of a fixed 120 ms
* perf(cli): check screenshots only the text contrast measures and skips samples with none
* perf(cli): seeks check fonts on the settle's second frame instead of waiting a third
* docs(cli): the contrast audit header says how a clipped screenshot is read
* docs(cli): cut the font-wait and contrast-audit comments to short blocks
* fix(cli): bound checks of non-finishing compositions
* fix(cli): account for cold startup in deadline witness
* fix(cli): wait for profile owners before test cleanup
* test(cli): separate portable deadline and Linux process witnesses
* test(cli): preserve per-launch browser profile ownership
* fix(cli): start check deadline after browser readiness
* feat(cli): scope check to a time range or a set of components
* fix(cli): keep ranged check gates truthful
Refuse ranged caption and frame gates when none of their seek times
falls inside the requested range. Make ranged motion reporting,
overlap timestamps, filtering tests, and scope documentation match.
* fix(cli): check warns when a video's browser-playable copy cannot be made
* fix(cli): big projects stop failing proxy prep, and a failed copy is reported once
* fix(cli): proxy failures name the real ffmpeg error, and a broken ffmpeg is reported once
* chore(studio-server): shorten the batch proxy resolver doc comment
* fix(core): a sub-composition timeline with a non-function duration no longer aborts the runtime
* fix(core): every registered-timeline duration read survives a partial or throwing entry
* fix(core): a runtime that fails to start reports its real error instead of a half-built page
* fix(core): a timeline post that throws after start-up is named too, and a later good post clears it
* fix(cli): check skips a custom clock whose time() returns null instead of reading it as stuck at 0
* fix(core): a recorded runtime start-up error stays set instead of being cleared by a later post
* fix(core): the runtime failure message says what to fix
* fix(core): the runtime failure is recorded even when the thrown value cannot be printed
* chore(core): format and regenerate the position-edits bundle for the startup error recorder
* fix(cli): keyframes and validate close the browser when the runtime fails to start
* fix(studio): text in Studio reads without em or en dashes
* fix(studio): the dash check also catches HTML entities, and peak reads Unavailable
* test(studio): the dash check test covers every entity form
* fix(studio): messages Studio shows from lint, sdk, core and the renderer drop dashes too
* fix(producer): the gate check stops early on the engine's new fallback wording
* fix(core): the preview loads the next scene's images before it starts
* fix(cli): check stops reporting image loads that its own scrubbing cancelled
* fix: preload only a coming clip's own images, and judge aborted images by whether they still decode
* fix(cli): count an aborted image as broken only when its decode fails
* style(cli): keep the aborted-image comment within the line limit
* test(cli): pin that a reload still pending at the cap is not reported
* test(cli): run the cap test on whatever timers are pending, and always restore real timers
* fix(cli): keep check --json stdout pure by sending proxy pre-resolve diagnostics to stderr
preResolveHostileMediaProxies() logged its summary and scan-failure lines via
console.info, which writes to stdout. With autoProxy on (the default), any
project containing HEVC/ProRes/AV1 media emitted that line ahead of the JSON
envelope from `check --json`, so JSON.parse(stdout) failed for consumers of
the machine-readable output.
Route both diagnostics through console.error (stderr), matching how the check
command already reports its own failures. Add regression tests asserting the
summary and scan-failure lines land on stderr and that stdout stays untouched.
* refactor(cli): shorten the pre-resolve stderr comments
* refactor(cli): drop the pre-resolve doc line the comment ratchet rejects
---------
Co-authored-by: Miguel Angel Simon Sierra <miguel.sierra@heygen.com>
Classify each <audio> element before Web Audio capture: same-origin,
CORS-opted-in, or a non-http(s) scheme stays on the primary
createMediaElementSource() path; cross-origin media without a
crossorigin opt-in withholds that call (the Web Audio spec makes such a
node output silence without throwing) and falls back to fetch +
decodeAudioData, preserving the FX graph whenever the server allows
CORS. Recheck the route at the transport's irreversible capture
boundary, and account for currentSrc, src, and <source> candidates the
same way the HTML resource-selection algorithm does.
Emit a stable preview diagnostic (`runtime_web_audio_bypass`) at media
discovery time, not only from playback scheduling, so `hyperframes
check` — which seeks but never plays — can surface it as a
`web_audio_bypass` finding. Diagnostics are suppressed during export
rendering, where the producer mixes audio offline and already applies
the FX chain. The existing non-unit-rate fail-closed rule stays scoped
to fx-chain/automation so this fix does not newly mute grouped or
above-unity tracks.
Takes over #3459 with the data-native-audio escape hatch removed per
review feedback: the automatic cross-origin detection already covers
the cases that mattered, so the extra per-element opt-in attribute,
its route-classifier branch, and its diagnostic path are dropped in
favor of a single automatic behavior.
Fixes#3458
Original-Author: desenmeng
Co-Authored-By: desenmeng <desenmeng@users.noreply.github.com>
Co-Authored-By: Miga <noreply@anthropic.com>
* feat(media): alpha-capable authoring proxies
Alpha sources were refused a proxy before the codec map ever asked whether the
browser could decode them, so a ProRes 4444 alpha file (which no browser
previews at all) rendered black forever, while an alpha WebM (which previews
fine) was already covered by the browser-safe check on the next line. The alpha
veto earned nothing and cost the one case that needed help.
Alpha is now a target-codec choice rather than a veto: alpha sources transcode
to VP9 + yuva420p in WebM, everything else keeps the existing H.264/MP4 path
byte for byte. Only files no browser can preview are proxied, which is the rule
the runtime already followed everywhere else.
WebM cannot carry AAC, so the VP9 path uses Opus and drops the MP4-only
faststart flag. PROXY_PARAMS_VERSION moves to v3 so clients stop serving the
previously cached proxies.
Safari does not decode VP9 alpha and still shows black for alpha sources, as it
does today: this is better on Chromium and Firefox and no worse anywhere.
* fix(media): infer proxy variant for rescue
* fix(media): preserve alpha proxy hardening after restack
* feat(studio-server): serve H.264 proxies from the preview route
Wires the codec manifest and the transcoder into the preview surface: the route
negotiates a proxy via a query param and serves it through the existing range
and ETag machinery, composition HTML carries a codec map for the runtime, and
hostile assets pre-warm so a first play does not wait on a cold transcode.
Exposes the three subpath exports the CLI surfaces consume upstack.
Drops the TEMP fallow entry added with the transcoder: it has real importers now.
* fix(studio-server): publish media proxy exports
* fix(parsers): scan HTML comments linearly
* feat(cli): let projects opt out of automatic proxying
Adds media.autoProxy to hyperframes.json plus --proxy/--no-proxy flags, and
forwards the resolved value into the studio and preview servers and the vite
adapter. Lands before the runtime slice that turns auto-proxying on, so the
switch exists before there is any behavior to switch off.
* fix(cli): align media config schema
* feat(core): swap undecodable video to its proxy at runtime
Adds the browser-side half: before first load the runtime consults the injected
codec map and swaps a hostile source to its proxy, and if a video still reports
zero decodable width it rescues it reactively. An HEVC file carrying AAC fires
no error event, so zero videoWidth, not the error event, is the reliable signal.
Audio elements and alpha sources are never proxied, render mode never proxies,
and each swap evicts the element's stale sync state and reports once.
This completes the loop: auto-proxying is live for preview and studio from here.
The opt-out (media.autoProxy, --no-proxy) shipped in the previous slice.
* feat(cli): serve proxies from play, present, and the static project server
Adds proxy negotiation to the CLI-side servers and gives play byte-range
serving it never had, so a swapped video can seek. The static project server
behind check, snapshot, compare and friends injects the codec map once, so all
of its callers inherit the behavior; snapshot forwards its own proxy flag.
* fix(cli): serve proxies for camera formats
* feat(cli): resolve proxies before check's timed browser phase
check pre-resolves hostile assets so a cold transcode cannot exhaust the
render-ready budget, and surfaces the runtime's proxy diagnostics as findings
so a swap is visible rather than silent.
* fix(cli): harden proxy pre-resolution
The pipeline already carried FrameCheckOptions; only the flag was
boolean, which meant a pipeline caller tuning severity or seek points
would have them silently dropped — the two sides only agreed because
today's caller happens to match the defaults. Bare --frame-check keeps
the defaults; the value form mirrors --caption-zone's grammar, freezing
the contract before a release pins it.
Port validate's per-media-element clip audit into check's session
(clip_media_fit findings): an intrinsic duration meaningfully shorter
than the data-duration slot silently shortens the slot at render time,
and neither lint nor the runtime listeners can see it. A linter crash
now reports as check_lint_failure instead of masquerading as a runtime
failure. Finding-crop capture failures stay non-gating but emit a
stderr note and a telemetry error event so rollouts can measure the
second-session failure rate.
Three filters keep the escalated contrast gate honest, each surfaced by
running check across the registry examples:
- data-layout-ignore (the layout audit's existing decorative opt-out)
now also excludes set-dressing text from the contrast audit — one
vocabulary for 'not copy a viewer must read'.
- Text that has (nearly) left the canvas is skipped: sampling a clamped
off-canvas box reads border pixels and produced the classic false
white-on-white (a cursor exiting the frame).
- Contrast failures follow the same persistence rule as layout findings:
observed at a single sample of a multi-sample sweep demotes to
warning; held failures gate.
Example fixes the sweep exposed: motion-blur's 21 deliberately-dim rail
labels are marked decorative (its vivid labels pass on their own);
nyt-graph's subtitle and source-note adopt the gate's suggested
compliant gray; decision-tree's declared duration drops 15s to 10s —
its content ends at ~9.5s and every render shipped a blank white tail.
Every duplicated decision gets one owner: rectToBbox lives in checkTypes
(was verbatim in pipeline and browser layers); the audit seek tuning is
one exported AUDIT_SEEK_OPTIONS consumed by check and the deprecated
inspect path; zoom padding/scale defaults export from the capture module
instead of re-literalized in three files; the optional run_id property
is built by one helper across all three telemetry events; check's
--max-transition-samples parsing reuses its own positiveInteger helper;
validate drops a leftover re-export and redundant explicit-default args
go away.
Tests: the contrast candidate round-trip gains a real integration
anchor (the actual browser script eval'd in-page, a wrapper asserting
finish receives the page-script bbox shape) replacing regex-over-source
as the primary guard; the redundant geometry source-golden and a
duplicated deprecation-envelope assertion are dropped.
snapshot --zoom <selector|x,y,w,h> + --zoom-scale (default 3) crops via
Puppeteer clip at raised deviceScaleFactor — density changes, layout
never does. Selector resolves per frame with 24px padding; no match is
a loud error, and a frame whose clamped region is a sliver (element
collapsed or animated off-canvas) is skipped with a stderr note rather
than written as a useless few-pixel image.
check --snapshots additionally writes finding-NN-<code>.png crops for
error findings with bboxes (cap 12, deterministic re-seek in a second
session) and draws labeled annotation boxes on overview frames via a
transient overlay injected only after audits complete. Skill reference
gains the zoom workflow: check reports a finding, zoom into it, fix,
re-check.
HYPERFRAMES_RUN_ID (trimmed, 128-char cap) attaches as run_id to the
generic cli_command / cli_command_result events, absent when unset, so
an orchestrator setting it per design element can group a verify loop's
invocations in analytics. check additionally emits one check_report
event per invocation (including lint-short-circuited and failing runs):
gate booleans, per-class error/warning counts, launch/seek/contrast
phase timings, sample counts, ok and exit code. Timings stay internal;
no command output changes.
Ports the EF bridge's captionZone and frameCheck semantics as opt-in
flags so the bespoke bridge can be retired: --caption-zone takes
fractional band geometry (x0;y0;x1;y1) with optional severity routing
and seek points, defaults matching the bridge (caption seek [1], frame
seek [0.5], 2px tolerance, 0.05 opacity floor, 4px minimum size, 0.95
full-frame exclusion, center-in-band comparison, tag|text dedup).
--frame-check adds media bounds detection (img/svg/video/canvas) the
always-on text canvas_overflow never covered, reusing overflowFor.
Breach floor: max(120px, 6% of min canvas dimension). Band math derives
from the composition's own canvas, portrait included. Both gates off by
default; plain check output unchanged.