T8mars/minimax-h3-prompt-skill-T8 / skills/direct-seedance-3d-animation
direct-seedance-3d-animation
VerifiedPlan and prompt a stylized 3D animated short for Seedance 2.0 with stable character identities, environment continuity, causal story beats, shot-specific performance, camera, sound, and clip handoffs. Use when a story needs one 4–15 second generation or a longer multi-clip plan; not for photoreal live action, a single still, or generic prompt polishing.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 2.5 KB |
| agents/openai.yaml | config | 267 B |
| references/summary.md | markdown | 750 B |
| references/template.md | markdown | 1.6 KB |
references/template.md
View on GitHub# Seedance 3D animation template
## Production lock
```text
Format: [ratio], [total duration], [single clip / multi-clip]
Character A: [identity anchors, silhouette, wardrobe, emotional baseline]
Environment: [layout anchors, light direction, palette, fixed objects]
Story chain: [want -> obstacle -> action -> consequence -> payoff]
```
## Per-clip card
```text
Clip purpose: [new story information]
Opening state: [subject pose, screen position, environment state]
Action chain: [anticipation -> committed action -> contact/consequence -> recovery]
Handoff: [last pose, gaze, object motion, camera direction, sound tail]
References: [图片N/视频N -> one role each]
```
## Prompt skeleton
```text
镜头1:[stable subject binding and composition]. [visible performance and object action]. [primary camera behavior]. [light/material/style]. [sound].
镜头2:[continue the causal action without resetting identity or geography]. [new consequence and reaction]. [coherent camera]. [sound/dialogue].
镜头3:[resolve the clip goal and hold a handoff-ready final state].
约束:角色身份、服装、比例、场景锚点和光线方向稳定;无多余角色、无肢体融合、无风格漂移。
```
Use fewer shots when fewer causal changes exist. For a continuous performance, replace the blocks with one paragraph.
## Cross-clip audit
- Identity, wardrobe, scale, handedness, prop ownership, and environment landmarks match.
- Camera direction and moving objects enter the next clip consistently.
- Dialogue and sound tails are not duplicated.
- No clip exceeds 15 seconds and no Seedance prompt contains shot timestamps or H3 syntax.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |