← Back to list
video-cut Checked
A skill from the GitHub repository zenstory-ai/video-recap-skills.
8.7Overall rating
Installation
🤖 Install via AI
Copy the prompt below and send it to your AI assistant (e.g. Claude Code) — it will follow the instructions and install automatically.
Install the "video-cut" skill by following the instructions at https://skill123.me/install/video-cut.
⌨️ Command line install
Run in your terminal — downloads and installs to ~/.claude/skills/.
curl -fsSL https://skill123.me/install/video-cut.sh | bash
📦 Download ZIP
Download the zip and extract it into your skills directory (e.g. ~/.claude/skills/), then restart your session.
⬇ Download v1.0.0 · 30 KBAbout this skill
Overview
A skill from the GitHub repository zenstory-ai/video-recap-skills.
Source
- Repo: https://github.com/zenstory-ai/video-recap-skills
- Path: skills/video-cut
Score breakdown
Trigger
6.5
Accurate pipeline-stage description with bilingual trigger words; user-invocable=false; no explicit When-Not in the description itself.
Structure
8.7
90-line SKILL.md over a compact 8-script (2.3k lines) toolkit; no references tier, so progressive disclosure is minimal but appropriate for a single-CLI stage.
Workflow
9.2
Disciplined executor design: validated plan contract, audio-integrity-first boundary snapping with named blockers (unsafe_clip_sentence_boundary, clamped_beats), scene-cut candidate triage taxonomy, deterministic caching rules, and an explicit capability-boundary section.
Content
9.0
Precise rules with runnable commands (cut.py flags, ffmpeg scene filter with threshold caveats); consistent terminology throughout.
Engineering
9.0
Valid frontmatter; documented python3 invocation without shebangs; no external deps beyond ffmpeg; paths consistent.
Security
10.0
Capability manifest matches declared scope: zero network calls, no credential access, no dynamic code execution, writes limited to work_dir, stale-artifact cleanup scoped to its own QC file.
Cut stage of the zenstory-ai video-recap pipeline: consumes an Agent-authored clip_plan.json, snaps boundaries to sentence/silence anchors (avoiding flash-frame scene-cut risks), and renders edited_source.mp4. Capability manifest (6.0): reads only the explicitly passed video, args and work_dir artifacts (clip_plan.json, speech_boundary_anchors.json, narration.json legacy path); side effects confined to work_dir outputs; network destinations: none — all subprocess calls are local ffmpeg/ffprobe list-args, no shell; credentials: none; persistence: none. Capabilities match declared scope.