On a parallel-capture disk-verify or streaming-drain breach, the outer
catch cleared probeSession without first closing the still-owned session,
orphaning the probe Chrome process precisely when the retry was recovering
from GPU/memory pressure. Introduce closeOrphanedProbeForRetry so both
retry catches close the session (with defensive .catch that logs on close
error) before releasing the reference, and cover it with a focused unit
test asserting closure-before-clear and the swallow-and-warn behaviour.
Addresses Magi's REQUEST_CHANGES on #2749; also closes Rames' sibling
concern at the streaming-retry path (renderOrchestrator.ts:3093).
— Via
A `<br>` is a void element but `resolveSingleChildTextTarget` treated a lone
`<br>` child as the element's text target. So `getOwnText(<h1>A<br>B</h1>)`
read the `<br>`'s (empty) textContent → `text: null`, which consumers surface
as "not editable", and `setOwnText` wrote into the `<br>`, corrupting it
(serialized as invalid `</br>`).
Exclude void elements from the single-child text target; read `<br>` as "\n";
and rebuild the text/`<br>` run from the newline-separated value on write,
reusing existing `<br>` nodes so their identity (data-hf-id) survives an
in-place edit.
## What
New cross-sample layout check `rotation_pivot_drift` — flags a rotating element that should spin **in place** but pivots about the **wrong point** (e.g. a wheel whose spokes use a hardcoded px `transformOrigin` instead of `svgOrigin`/`%`, so they swing off-center while every existing check still passes).
Motivating prod case: a portrait ad's spoked-wheel whose `#spokes` rotated about `transformOrigin:"250px 250px"` in a resized 460px container — spokes detached from the hub, shipped clean because no rule inspects rotation.
## How
- `layout-audit.browser.js`: `window.__hyperframesRotationSample()` reports each visible transformed element's bbox center + decoded rotation angle per layout sample. Skips `[data-layout-allow-orbit]`.
- `checkPipeline.ts`: accumulates samples across the seek grid; `detectRotationPivotDrift()` (modeled on `detectSweepStatic`) flags an element that (a) actually spins (angle spread > 20° over ≥3 samples), (b) is size-stable (bbox width ratio ≤ 1.6), and (c) whose bbox **center** drifts > `max(10% of its size, 2% of min viewport dim)`. Emits `warning`; not persistence-tiered (not demoted to info).
## FP guards
Real rotation required, ≥3 samples, size stability, `data-layout-allow-orbit` exemption, min area ~2500px². Center-drift (not bbox size) is the discriminator, so a correctly-centered spinner reads drift ≈ 0.
## Validation (`check --json`)
| Fixture | Expected | Result |
|---|---|---|
| off-transformOrigin spoked wheel | fire | **fired — 109px drift on `#spokes`** |
| non-spinning comps (node diagram, device tree) | clean | clean, no FP |
| correctly-centered spinner (`svgOrigin`) | clean | clean (spins 162°, drift 0) |
| `data-layout-allow-orbit` off-origin spinner | clean | clean (exempt) |
| off-`svgOrigin` control, no opt-out | fire | fired — 251px drift |
No false positives. `tsc --noEmit` clean, `oxlint` clean, `check.test.ts` + `layout-audit.browser.test.ts` = 112/112 pass.
## Note
`ROTATION_MAX_SIZE_RATIO` is 1.6 (not 1.3): a rotating anisotropic shape's axis-aligned bbox inherently oscillates (8-spoke star ~1.32×, square 1.41×), so a tighter ratio rejects legitimate targets. Center-drift stays the real discriminator; thin swinging bars are excluded.
Follow-up: a `detectRotationPivotDrift` unit test via the fake driver's `collectRotationSample` (mirroring the sweep_static tests).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
The pre-commit build/lint/typecheck PreToolUse hook emitted {continue:false,
stopReason}, which halts the agent's turn — the dev had to type 'continue'
after every failed gate. Switch to a PreToolUse deny decision so the failing
commit is still blocked but the failure text is fed back to the agent as tool
feedback, letting it fix and retry without a manual nudge.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Chrome 112+ / Firefox 117+ / Safari 16.5+ support native CSS Nesting.
A nested rule like '.title { … }' inside '[data-composition-id="intro"]
{ … }' resolves at match time to '<parent> .title' via the implicit
'&' prefix.
'scopeCssToComposition' walks every rule via 'root.walkRules' and re-
scopes selectors, but it did so for nested rules too — producing
'[…scope…] .title' inside '[…scope…] { … }', which nesting then
prepends AGAIN to '[…scope…] […scope…] .title'. Since the composition
root only appears once in the DOM, the doubly-scoped selector never
matches — the nested rule appears 'just ignored' as the reporter
described (#2721).
Reproduced on 0.7.66 with the reporter's exact composition. Fix: add
'isNestedInsideAnotherRule' predicate — mirrors the existing
'isInsideGlobalAtRule' — and skip nested rules in the walkRules
callback. Top-level rules still get scoped; their nested descendants
inherit scope naturally via CSS Nesting at match time.
Added two focused tests:
- 'preserves nested-rule selectors so CSS Nesting inheritance works
(#2721)' — asserts nested '.title' and 'h2' selectors stay verbatim
while top-level rules keep scoping.
- 'preserves deeply-nested CSS Nesting rules (#2721)' — same rule at
depth 3.
All 37 existing scopeCssToComposition tests still pass.
Fixes#2721.
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
#2606 taught the video-coverage gate that a non-looping short video holds
its final decoded frame across the tail, so the delivered source frames
are enough to cover the authored slot. But that fix gated the credit on
'!video.loop' — a looping short video was still measured as
unique-source-frames / slot-frames and aborted at ratio << threshold.
Reproduced on 0.7.64 with the reporter's exact composition (3s clip in a
10s slot, loop attribute): render aborts with 'captured 90 of expected
300 frames (coverage 30.0%)'. Same source without loop renders clean via
#2606's freeze credit. This is the mainline 'loop a short clip to fill a
longer scene' case, the reason loop exists.
Fix: extend #2606's source-credit to loops symmetrically — the delivered
set (all N source frames) covers every repeat within the slot, so
expectedFrames = min(slotFrames, sourceFrames) for both hold and loop.
Fail-loud preserved for a genuinely-broken loop (extractor truncated
below its own source): a 60/90 delivery still aborts at 66.7% < 95%
because the delivered set no longer covers the full source period the
loop reuses. Missing extractions still require the full slot.
Test updates:
- 'still requires the full authored slot for looping clips' locked in
the buggy behavior; replaced with 'credits a looping short clip
against the source portion' which asserts the correct 90/90/1.0.
- Added 'still fails when a looping clip's source extraction is
truncated' as the new fail-loud floor.
Fixes#2665. Regression window: 0.7.60 (#2606's original ship)
through 0.7.67 (current).
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
* docs(skills): make captions non-optional in changelog-video
The Jul 13-20 build shipped without captions because the pipeline had two soft
failure modes:
- Empty vo-words.json from the TTS provider was silently accepted, so
align-captions.mjs had nothing to align. captions.json was never produced.
- Step 5 tolerated the master-skeleton's LINES array staying empty. The build
lint-checked and validated clean, but the caption rail rendered no text.
This adds three hard gates on top of the existing scaffold:
- Step 4 adds a whisper forced-alignment fallback so a missing TTS timestamp
payload no longer breaks the caption pipeline. Whisper only supplies timings;
captions still use the DISPLAY layer from script-tokens.json.
- Step 5 flags an empty LINES array as a shipped bug and shows the exact
IIFE-input shape it expects.
- Step 6 adds gate 5: sample 3-4 frames across the VO window and confirm visible
caption text on each. If any spoken window renders no caption, the build is
red.
Anti-patterns table gets two rows covering the empty-LINES and
missing-word-timings failure modes. .claude and .agents mirrors stay
byte-identical.
Follow-up to #2669 (pre-build gate); paired with the v5 rebuild that
retrofitted captions onto the shipped Jul 13-20 video.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* docs(skills): bump changelog caption size to 32px for phone-screen legibility
25px reads too small at 1080² when the video is viewed on a phone without
zoom. Bumped to 32px (font-size 25→32, top 1002→990, height 40→52, alpha
.92→.94) so captions carry across small-screen playback. Change is CSS-only
on the master-skeleton; existing per-scene chrome and layouts are untouched.
Retrofitted the Jul 13-20 v6 build with the same values (identical CSS
block) so the shipped video and the future default match. .claude and
.agents mirrors stay byte-identical.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* docs(skills): sync caption position refs to top: 990 across SKILL.md + build-spec
Miga caught: the master-skeleton bump to top: 990 / 32px / height: 52 was
not reflected in the SKILL.md prose or build-spec.md. All three files now
name the same values so the doc, gate, and code stay consistent. Also
clarifies build-spec.md's caption-rail line to point at SKILL.md step 5 as
the source of the LINES-populated requirement.
.claude and .agents mirrors byte-identical.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* docs(skills): sync script-voice.md caption rail to top: 990 / 32px
Third and final stale reference to top: 1002 / 25px — script-voice.md was
still stating the pre-bump values. Now matches master-skeleton.html,
build-spec.md, and SKILL.md. Also bumps the 'keep critical small text out'
guidance from ~80px to ~100px to reflect the slightly taller caption rail.
Full-directory sweep confirms zero remaining occurrences of 1002 or 25px
across .claude/ and .agents/ skill mirrors.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
* docs(changelog): weekly digest 2026-07-13–2026-07-20
* docs(changelog): embed the weekly video in the Jul 13-20 digest
Adds the Jul 13-20 changelog video (produced via the changelog-video
skill, hosted at static.heygen.ai) at the top of the Update block using
Mintlify's <Frame> component with a native <video controls>.
Video URL: https://static.heygen.ai/hyperframes/changelog-videos/weekly-changelog-jul13-20.mp4
Slack thread: C0ACCNHLG3U · 1784565857.520599
Linear: VA-1859
## Problem
For a `fit_to_scene` B-roll where the composition's intrinsic timeline (e.g. `data-duration=1.0s` → 30 frames) is shorter than the scene it fills (e.g. 4.8s narration), the producer renders only the intrinsic 30 frames and the downstream compositor frame-holds/PTS-stretches that fixed clip to the scene length. Spreading 30 unique frames over 4.8s starves motion to ~6 effective fps → a visibly choppy result. Root cause: the producer welds one `composition.duration` to both the frame count and the 1:1 seek mapping, with no notion of a target output length.
## Fix
Add optional `renderStretch: number` (default `1.0` = no-op), `renderStretch = intrinsic / target`:
- **Frame count** comes from the target: `outputDuration = intrinsic / renderStretch`, `totalFrames = outputDuration × fps` (`probeStage.ts`). `composition.duration` stays intrinsic (drives video/audio windows).
- **Per-frame seek** is scaled: `time = (frameIndex / fps) × renderStretch`, so the N output frames map across `[0, intrinsic]` — a fresh frame per output frame.
All seek sites go through a single shared `outputFrameToTimelineSeconds(frameIndex, fps, renderStretch)` helper (`core.types.ts`), consumed by every capture path so none can silently diverge:
- parallel (`parallelCoordinator.ts`), `sdr_streaming` (`captureStreamingStage.ts` ×3), `sdr_disk` (`captureStage.ts`), HDR loops.
- DrawElement + static self-verify (`frameCapture.ts`) — ground-truth seek uses the same mapping, so PSNR compares like-for-like (no spurious verification failure on stretched comps).
- Distributed path: `renderStretch` threaded through `DistributedRenderConfig` → chunk workers, and **folded into the plan hash only when `!= 1`** so a pre-stretch cached plan is never reused.
With `renderStretch = 1` (or omitted → `?? 1`): every seek is `×1.0` (IEEE-754 identity), frame counts unchanged, and the plan hash is byte-identical — a provable no-op. `player.ts` absolute-seek is untouched.
## Verify
- typecheck (core + engine + producer): pass. lint/format/fallow/commitlint: pass. `planHash` + `renderRequest` unit suites: pass.
- Adversarial self-review found + fixed three capture-path gaps (streaming, self-verify, distributed) before this revision.
- **Not yet runtime-verified** on a real render — needs a fit_to_scene render at `renderStretch < 1` confirming N distinct frames over the target length (draft until then).
Paired with experiment-framework#42766, which computes and forwards `renderStretch = hf intrinsic / scene duration`.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
* fix(studio): play preview audio at speeds above 1x
The player force-muted preview audio whenever playback rate exceeded 1x
and disabled the mute button (and the M shortcut) in that state. Users
changing speed to 2x heard nothing, and clicking unmute did nothing.
Media elements play audio fine at any rate (pitch preserved by default),
so remove the special case: audio now follows only the user's mute
toggle at every speed. Drops the shouldMutePreviewAudio helper and the
audioAutoMuted UI/keyboard gating that existed solely for it.
* test(studio): update seek audio test for unmuted-above-1x behavior
useTimelinePlayer.seek.test.ts still asserted the old force-mute: at 2x
with the user unmuted it expected a set-muted{muted:true} message. With
audio now following only the user's toggle, the preview receives
set-muted{muted:false}. Update the assertion and title to match.
* feat(skills): c2v mining pass over animation blueprints and rules
Compacts ~45 existing animation rules/blueprints into tighter recipe form
(net -3.4k lines) and adds 17 mined from the c2v corpus:
- 7 blueprints: agent-progress-theater, camera-journey, fixed-anchor-cycle,
panel-edit-live-sync, prompt-type-submit-generate,
transcript-scroll-artifact-reveal, zoom-out-workspace-reveal
- 10 rules: 3d-camera-flight, anchored-layout-expand, chart-scrub-readout,
chromatic-glitch, control-target-sync, cursor-drag, gradient-text-sweep,
multi-cursor-choreography, particle-burst, theme-crossfade-morph
Both indexes updated.
* feat(skills): sync product-launch script bank with mined blueprint roles
The role->blueprint script bank in product-launch-video/story-design.md is
kept 1:1 with blueprints-index role declarations, which the c2v mining pass
expanded. Adds the 25 missing entries (script-shape descriptor + example
lines + pattern): 13 for the 7 new blueprints, 12 for role widenings on 7
existing ones (cursor-ui-demo, dataviz-countup, titlecard-reveal, et al.),
and states the 1:1 sync contract in the bank's intro.
* docs(skills): cover constellation-hub scatter-drift variant in the script bank
Review follow-up on #2680: the SOCIAL_PROOF constellation-hub entry
patterned only the orbit shape; the c2v pass added a scatter-drift
end-card variant with the opposite geometry (no hub, no ring). Adds an
example line and extends the pattern so a scatter-drift beat's VO isn't
steered toward the orbit shape.
The reproduction packet template in hyperframes-cli previously said
"Include the rerunnable command and working directory" and shipped a
`REPRO COMMAND: cd <project path> && ...` skeleton. Agents faithfully
followed both, so user home-directory prefixes (`/home/<user>/projects/...`,
`/Users/<user>/Documents/...`) have been landing verbatim in the public
CLI feedback channel — leaking user + machine identity that maintainers
don't need to reproduce a bug.
Fix, docs-only:
- SKILL.md: change "rerunnable command and working directory" to
"rerunnable command (relative to the project directory)" plus an
explicit note that feedback is public and absolute paths must not be
pasted. Add a matching redaction rule for EXACT ERROR stack traces
(keep basename+line, drop leading directory).
- references/preview-render.md: replace the `cd <project path>` skeleton
with a bare `<HF_*/PRODUCER_* env> npx hyperframes <exact command>`
template + inline comment reminding to run from the project directory
without pasting absolute paths. Rewrite the "Preserve paths / redact
secrets" line to lead with the anonymization rule and give concrete
before/after examples (`./renders/out.mp4` vs
`/Users/<user>/Documents/…/out.mp4`).
The COMPOSITION_STRUCTURE block was already privacy-preserving (counts
+ presence flags only, "no file paths, no src URLs, no user text") — no
change needed there.
skills-manifest.json regenerated for the hyperframes-cli hash.
Signed-off-by: Via <noreply@anthropic.com>
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Adds an explicit Step 0 above the pipeline that spells out the concrete file
copies (fonts, bgm, ffmpeg-encoded bg-pattern, master-skeleton.html) before
any composition HTML is written. Extends the anti-patterns table with the
three failure modes James caught on the Jul 13-20 weekly run: starting from
a prior video's index.html, hand-crafting @font-face / WebGL shader / custom
BGM, and shipping without a CloudFront invalidation after the S3 replace.
Why: the previous SKILL.md said "follow references/build-spec.md exactly" in
step 5, but that lets an agent read the build-spec after already writing
composition HTML on the wrong scaffold. Ships of the Jul 13-20 video went
through three off-brand iterations before I finally started from
master-skeleton.html (Jake called v3 "trash" — the skill's assets, fonts,
and scaffold ARE the brand, not the SKILL.md prompt on its own).
The three added anti-patterns are the concrete traps: agents will keep
starting from prior-project templates unless the skill explicitly bans it.
Mirrors to .agents/skills/ (Codex CLI) and .claude/skills/ (Claude Code)
byte-identical, per the skill-mirror gate.
The Claude Design Send-to guide rendered at its URL but was an orphan — absent from the
docs nav and llms.txt, so it never surfaced as a fetchable search RESULT (only as body
text in another page's excerpt, which Claude Design's web_fetch rejects). Add it to the
Guides nav (→ llms.txt index) next to claude-design, and give it frontmatter title +
description for a clean entry. This makes the full guide discoverable + fetchable via
search, the reliable path since a URL named only in a tool description is not fetchable.
Vance Ingalls