| description | Prepare deterministic transcript artifacts for pi from a transcript output directory |
|---|
Prepare transcript analysis artifacts for: $@
Workflow:
- Identify the transcript output directory produced by
scripts/transcribe_video.py. - Derive a stable episode slug from the argument or transcript directory name. If the slug is ambiguous, ask one brief clarifying question.
- Run:
uv run python scripts/prepare_transcript_analysis.py <transcript-dir> --output-dir dist/analysis/<slug> --slug <slug> --plain-output-names - Keep outputs under
dist/analysis/<slug>/unless the user or profile requests a different layout. - After the script finishes, report the exact output paths that were written.
Rules:
- Do not do editorial analysis in this step.
- Do not invent missing transcript inputs.
- Prefer reusing existing prepared artifacts when they already match the requested transcript and slug.