菜单
Author the per-lesson sound manifest (audio-cues.json) — bed mood, intro sting, outro resolve, and the SFX→beat map. Wave 2c of the lesson pipeline, parallel with audio/captions. SEMANTIC only — declares moods and sound keys, never frames.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
| name | lesson-sound-design |
| description | Author the per-lesson sound manifest (audio-cues.json) — bed mood, intro sting, outro resolve, and the SFX→beat map. Wave 2c of the lesson pipeline, parallel with audio/captions. SEMANTIC only — declares moods and sound keys, never frames. |
Decide what the lesson SOUNDS like, not when sounds fire. You author one small JSON file; the composer (Wave 4) maps your semantic beats onto its own motion frames, and the sound-asset lane (Wave 3c) guarantees the assets exist. You touch no code and no WAVs.
Run in parallel with 2b (audio/captions). You depend on visual-design.md (Wave 2a) + pedagogy.md (Wave 0), NOT on voice. Music+SFX is a SECOND audio track that consumes the reconciled timeline — it never drives cue length and is never frozen at Wave 3a.
Research basis: research/music-sound-design.md (mix), research/music-sound-palette-2026-05-29.md (palette + SFX + the tone-language guard). Integration + schema: docs/proposals/sound-layer-integration.md. Code surface: CAPABILITIES.md#lesson-music-bed, #lesson-sfx-layer.
lesson-data/<id>/brief.md — topic → bed mood + whether this is a tone/pinyin lesson.lesson-data/<id>/pedagogy.md — the discovery/reward beats → where the single ta-da lands.lesson-data/<id>/visual-design.md — per-cue SFX-worthy tags (entrance / count / transition / success).lesson-data/<id>/audio-cues.json — schema = src/lesson-media/audioCuesTypes.ts (AudioCues).lesson-data/<id>/_logs/sound-design.md — inputs read, decisions, the bed/SFX keys you chose and why.Pick ONE bed key from the curated library by mood — never invent a track per lesson:
| Lesson kind | bed | toneSafe |
|---|---|---|
| Number / math (default) | math-calm-68-cmaj | false |
| Literacy / phonics | literacy-playful-76 | false |
| Pinyin / tone-teaching | tone-safe-pad | true |
| (outro/celebration moments use the envelope's outro resolve, not a separate bed) | — | — |
The tone-language guard (the rule unique to us). When the child must hear a spoken lexical tone (pinyin lessons), a melodic bed pitch-contour competes with the tone. Set toneSafe: true → the bed is a flat pad/drone and NO melodic motif (sting/pentatonic accent) may play under narration. A pentatonic dizi/guzheng accent is welcome, but only in the intro or a narration GAP, never under tone-teaching words.
Palette discipline (from research — do not re-litigate per lesson): beds are calm, major/neutral, 60–80 BPM, no vocals. Western soft piano + warm pad is the safe default; Mandarin pentatonic instruments are a sparing ACCENT (intro/success/gaps), not the bed. Default to the existing library; if a needed mood is genuinely absent, NAME it for the Wave 3c sound-asset lane — do not describe a bespoke score.
intro.sting: optional one-shot over the topic-intro card. A mandarin-accent sting is allowed for Mandarin topics (it sits over the intro card, before narration).outro.resolve: true: the bed rises to full as the last narration ends. Default on.Map only MOTIVATED beats, using the fixed vocabulary (pop / chime / whoosh / tick / ta-da). Each row is { cue, event, sound, perStep?, risingPitch? } — semantic; the composer computes the frame.
event | Fires on (composer's motion) | Typical sound |
|---|---|---|
popin | a <PopIn> entrance | pop |
count | each counted step (set perStep: true) | tick |
transition | section/scene change, <ShineSweep>, SectionHandoff | whoosh |
reward | a discovery/success beat | chime (small) or ta-da (the win) |
ta-da at most ONCE per lesson, on the single success beat.risingPitch: true (with perStep) ascends pitch across a counted set — auditory magnitude. This is evidence-INFORMED, not proven for counting specifically; use it, flag it [ASSUMED] in your log, and expect Wave 6 to sanity-check it.audio-cues.json (frames are the composer's; the schema has no frame field).toneSafe.visual-design.md.audio-cues.json + the bed/toneSafe choice with one-line reason.[ASSUMED] choice (e.g. risingPitch) flagged for Wave 6.Viewer-first discipline for early-childhood Remotion lesson scenes. Before any choreography or JSX, the subagent must inhabit the kid's eye — measure how the smallest teaching mark and the smallest label actually render at the composition size, declare disjoint visual zones, refuse decoration. Pairs with visual-discipline (contract mechanics) and early-childhood-visual-taste (color/tone).
Use when writing per-cue teacher narration and captions for a Remotion lesson. The narration is COMMENTARY ON THE VISUAL — Wave 2a's per-cue motion budget is the input, and you write narration to fit each cue's budget. No invented hold timings.
Concrete visual discipline for Remotion lesson scenes — Visual Contract before choreography, one-metaphor rule, container earns existence, occupancy budget, semantic groups dominate visual groups, render-and-self-critique loop, iteration economy. Pairs with kids-eye (viewer-first) and early-childhood-visual-taste (color/tone).
Compose the Remotion lesson scene from approved upstream artifacts (Visual Contract, audio-captions intent, ASR-aligned cue boundaries, sketch-overlay schedule). Wave 4 of the lesson pipeline. NEVER hardcodes frame numbers; every timing derives from cue boundaries.
Visual taste — palette, tone, motion vocabulary — for early-childhood Remotion math lessons. Companion to kids-eye (viewer-first) and visual-discipline (Contract mechanics). Use during Wave 2 visual-design to set the lesson's visual language AFTER the kids-eye measurement block and BEFORE the Visual Contract.
Specify teacher-mark overlays (Excalidraw-style hand-drawn ink) for a Remotion lesson scene. Wave 4 (parallel with composer). Marks are specified in CUE-RELATIVE frames, never master-timeline absolutes. Restraint is the rule.