screenkite-clean-cut

Remove silence gaps and filler words (um, uh, ah, er, hmm) from a ScreenKite recording in a single pass — one transcript, one dry-run, one cut.

Prerequisites

• screenkite CLI at /usr/local/bin/screenkite-alpha (or screenkite).
• ELEVEN_LABS_API_KEY (underscored) in a .env anywhere up from the bundle, or exported in the shell environment.
• ffmpeg on PATH.
• python3 + requests — use uv run --with requests to avoid install issues.

Phase 1 — Open project

'/usr/local/bin/screenkite-alpha' agent project open \
  --path '<absolute-path-to.skbundle>' --json