tighten-dialogue

I'd like you to add a skill to Wideframe called "tighten-dialogue".

What it should do:
Remove pauses and filler words down to a max gap, keeping audio and video in sync.

Before you run anything, ask me to confirm:
- Maximum allowed silence gap (seconds or frames)?
- Remove filler words (um, uh, like) too?
- When a line has multiple takes, prefer which (last, cleanest)?
- Back up the original sequence first?

Then do the work:
1. Transcribe with word-level timing using media-transcript-get-batch.
2. Measure the loudness envelope with media-audio-envelope to confirm the silence spans.
3. Mark silences over the max gap and any filler words.
4. Ripple-delete the marked ranges with sequence-clips-remove and sequence-clips-update.
5. Verify audio and video stay aligned across each cut.

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:
- No silence gap in the output exceeds the stated maximum.
- Flagged filler words are absent from the cut.
- Audio and video remain in sync across every cut (spot-checkable).
- A removed-ranges log lists every cut with timecodes.

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 "tighten-dialogue" so I can run it again by saying /tighten-dialogue.