Rn Motion Replica
Analyze a public or local reference video, isolate a time range, deconstruct its...
Install
npx skills add Pluviobyte/rnskill --skill rn-motion-replica -g -a claude-code -y- Repository
- Pluviobyte/rnskill1.7k
- Category
- Media
- Updated
- 1mo ago
- License
- See repository
- Context cost
- Mediumabout 1,592 tokens
- Runs scripts
- Yes, review before use
What it does
Analyze a public or local reference video, isolate a time range, deconstruct its layout, motion, timing, and transitions, then build an original verified motion replica with reusable templates and final MP4 QC. Use when the user says 复刻这个视频效果, 参考这个视频做同款动效, 拆解前 N 秒, 还原方框/连线/卡片动画, or provides a video or URL as a motion reference rather than asking to copy its footage or script.
RN Motion Replica
Turn a reference clip into an original motion study that preserves the visual grammar and timing while replacing the source's copy, presenter, footage, and unique assets.
This Skill targets an editable original adaptation. Route any pixel-exact,
render-frame-exact, or encoded-frame-aligned claim through rn-replica-qc and
its asset, runtime, and delivery gates.
Source and rights boundary
- Work only from a public link or local file the user is authorized to use.
- Do not bypass private-account controls, DRM, paywalls, or deleted media.
- Keep the source URL, title, downloaded media, transcript, and source screenshots in a private working directory.
- Do not ship the source media inside the finished project or Skill package.
- Recreate motion structure, layout logic, pacing, and interaction language. Write new copy and use original or generic assets.
Workflow
1. Lock the study range
Identify the exact time range, the requested effects, the delivery aspect ratio, and whether the user wants a study preview or a full production asset. When the request is already specific, proceed without another intake round.
Default the canvas and fps to the reference unless the user or destination requires another format.
2. Acquire and verify the reference
Download the public reference with the workspace's trusted downloader or yt-dlp. Verify it with ffprobe; an HTTP success is not enough.
Require:
- a playable video stream
- non-zero duration
- known dimensions and frame rate
- an untouched source file in the private reference directory
3. Generate visual evidence
Run the bundled analyzer on the selected range:
python3 scripts/analyze_reference.py <reference.mp4> \
--start 0 --duration 35 --out <private-analysis-dir>
Inspect all generated contact sheets and scene frames. Do not infer choreography from one thumbnail.
Record five layers of evidence:


