feat(skills): product-launch-video skill + consolidate motion knowledge into hyperframes-animation (#1745)

* feat(skills): product-launch-video + consolidate motion knowledge into hyperframes-animation

- Add the product-launch-video skill: shot-sequence architecture where each
  visual frame is a time-coded shot sequence picked from a blueprint menu and
  paced to the voiceover (anti-PowerPoint). Includes the frame-worker sub-agent,
  story/visual/motion-design references, and audio/captions/transitions/
  stage-assets/assemble-index scripts.
- Consolidate motion knowledge in hyperframes-animation as the single source of
  truth: promote the updated atomic rules (31 -> 36) and rename product-launch-
  video's archetypes into hyperframes-animation blueprints (13 -> 15, replacing
  the old set). product-launch-video, faceless-explainer, and pr-to-video now
  reference them via ../hyperframes-animation/{rules-index,blueprints-index}.md
  and the rules/blueprints dirs. Fixes the discrete-text-sequence broken links;
  blueprints no longer ship per-id runnable examples, so example references in
  the consumers were dropped.
- Default HeyGen TTS voice to Marcia (deterministic; was the API's first English
  voice, which drifts on catalog re-sort). Override with --voice.
- assemble-index pre-assembly frame guards: auto-repair a sub-comp root missing
  canvas dims; hard-fail on <video>/<audio> inside a sub-comp; hard-fail on a
  timed non-root element missing class="clip" or overlapping same-track clips.
- Lint/CLI: lint media inside sub-compositions as an error; stop false-positive
  caption layout/lint findings; contrast/layout-audit skip elements hidden by an
  invisible ancestor.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(skills): clear CodeQL alerts in assemble-index.mjs

- script/style blanking regex now matches closing tags with trailing
  whitespace (</script >, </style >) — js/bad-tag-filter (high).
- drop the existsSync precheck before reading/repairing a frame file; read
  directly and handle ENOENT, removing the check->write TOCTOU window —
  js/file-system-race (high).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

skills/hyperframes-animation/blueprints/video-text-pivot.md

@@ -0,0 +1,30 @@
+# video-text-pivot — Video → Text Pivot
+
+**intent**: A product video holds center and claims attention, then slides aside to hand its weight to a hero stat in the space it vacates, then both clear and kinetic text types into the center — accent words carrying the meaning the video used to carry — sealed by a gradient pill. The arc is "show → yield → pivot → stamp," and each handoff pairs an exit with a same-anchor entrance so two beats read, not four.
+
+**roles served**
+
+- Product_Intro (from `metric-video-text-pivot`): when the open is "see the feature" then "see the impact" and the `[product video]` must stay visible through the stat reveal — it slides, it doesn't cut.
+- Key_Feature: a feature clip that yields to a frame-filling metric and a typographic impact line.
+
+**duration**: 6–8s
+
+**shot structure** (a `[bg]` canvas; one `[product video]` as a real muted `.mp4` clip, a hero stat, then kinetic text — each pair shares a screen anchor so the handoff reads as a weight-transfer)
+
+- **Scene 1 (0.0–~1.6s) — the video shows.** The `[product video]` lands centered on a smooth scale-up and breathes (a small y-bob), claiming full attention. Camera static.
+- **Scene 2 (~1.6–3.2s) — yield + stat (signature move).** The video SLIDES aside (x + scale down) **into the very space** the `[hero stat]` now fills as the stat pops in with 3D-depth type — one weight-transfer reading as a single event, not two. The stat breathes within this window.
+- **Scene 3 (~3.2–5.0s) — pivot to text.** Both video and stat clear out and kinetic `[impact text]` TYPES into the vacated center, character by character; its `[accent words]` carry the meaning the video used to carry.
+- **Scene 4 (~5.0–end) — stamp.** A gradient `[pill]` snaps shut around the closing line (`scaleX` 0→1), its glow halo resolving a beat behind so the silhouette reads before the bloom — sealing the statement as one graphic. Holds.
+
+**motion vocabulary**: video scale-in + small breath; weight-transfer slide (video x + scale-down handing off to the stat at the same anchor); 3D-depth stat type; character-stream typing; gradient pill scaleX-snap; glow-halo bloom trailing the silhouette.
+
+**rule mapping**
+
+- video entrance (smooth) and the weight-transfer slide → `gsap-effects` (scale/opacity then x + scale on a long-tail `power3`); the video itself is a muted `<video class="clip">` direct child of the root
+- hero stat's frame-filling 3D type → `3d-text-depth-layers` (static-depth variation — layers built at setup, no cascade fighting the entry)
+- the same-anchor video-exit ↔ stat-entry handoff (if treated as a morph) → `scale-swap-transition` (shared center)
+- character-by-character impact typing through segmented spans → `dynamic-content-sequencing` (clean character stream) or `discrete-text-sequence`
+- pill `scaleX` snap + trailing glow halo → `gsap-effects` (scaleX) + `ambient-glow-bloom` (the halo, resolving a beat behind)
+- video / stat breath within their windows → `sine-wave-loop` (low-amplitude register — subtle jitter, gated to each element's window, never a forever loop)
+
+**camera modifier**: camera-static — all motion is element-space (the video translates), so the "pivot" is the elements moving, not a camera.