rename-clips

I'd like you to add a skill to Wideframe called "rename-clips".

What it should do:
Rename camera-original files based on what’s actually in them, on your naming convention.

Before you run anything, ask me to confirm:
- Naming pattern + field order (e.g. {date}_{project}_{subject}_{shot}_{take})
- Where each field comes from: content analysis, the slate, or filename/EXIF
- Date format + casing
- Rename in place vs. copy to a new folder (default: copy)

Then do the work:
1. Glob video files under $1. Pull duration/codec/created-date via ffprobe.
2. Run caltools media-analyze-content-batch to get shot type, on-screen subject, and scene description per file.
3. For dialogue clips, pull the first lines via caltools media-transcript-get-batch to confirm the speaker/slate.
4. Compose each new name strictly from the confirmed pattern. Leave unknown fields blank rather than guessing.
5. Write a before->after manifest (manifest.csv) next to the output. Apply the rename only after the manifest is written.

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 input maps to exactly one output, no collisions (auto-suffix _02, _03 on dupes).
- manifest.csv lets the user reverse every change.
- Report the before->after table and flag any file the agent could not confidently name.

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 "rename-clips" so I can run it again by saying /rename-clips.