ADD THIS SKILL
paste into a new chat
Copy this into a blank Wideframe chat. It tells the agent to build the skill, what to confirm with you first, the steps to follow, and how to verify the result.
PROMPT
I'd like you to add a skill to Wideframe called "split-by-silence".
What it should do:
Cut one long recording into clean, trimmed takes at the pauses.
Before you run anything, ask me to confirm:
- Silence threshold (dB) + minimum silence duration (s)
- Min / max clip length
- Handle (padding) before/after each clip
- Output: separate media files or cuts in one sequence
Then do the work:
1. Analyze the loudness envelope to find silence spans below the threshold.
2. Keep only silences longer than the minimum duration as split points.
3. Enforce min/max clip length (merge slivers, split overlong blocks at the quietest interior point).
4. Cut with the confirmed handles via ffmpeg (files) or sequence-clips-update (sequence).
Do not grade your own work. When the output is ready, hand it to a separate verification subagent that has its own context window. Give it the output, the choices I confirmed, and the pass criteria below. It reviews the quality, writes specific feedback, and returns PASS or FAIL.
Pass criteria:
- Every cut sits inside detected silence, never inside speech.
- All clips respect min/max bounds.
- Total output runtime reconciles with the source (report the delta).
If it returns FAIL, apply the feedback, redo the affected work, and verify again with a new subagent. Repeat until it returns PASS, then show me the result.
Once it works, save it as a reusable skill named "split-by-silence" so I can run it again by saying /split-by-silence.RUN IT AGAIN
Once it’s saved, run it on the next project with /split-by-silence, or just say:
“Split this 3-hour stream into takes wherever there’s 2+ seconds of silence, half-second handles.”Run it on your own footage.
7-day free trial. Requires Apple Silicon.
RELATED SKILLS