- Implemented the MVP job pipeline: ingest, transcribe, plan, write, format, package.
- Added SQLite-backed job storage, stage tracking, jobs API, and CLI execution.
- Added deterministic fallback content so a job can complete without API keys, captions, ffmpeg, or pandoc.
- Set up backend/frontend dependency locks and fixed test/build compatibility issues.
uv run pytest-> 10 passeduv run ruff check src tests-> passeduv run mypy src-> passedcorepack yarn build-> passedcorepack yarn test --run --passWithNoTests-> passed; no frontend test files exist yet
- API job submission currently runs the MVP pipeline inline and returns the completed job.
- Real transcript fetching is disabled by default; set
config.fetch_transcript=trueto attempt YouTube transcript fetch. - EPUB/PDF/cover generation remain future work; MVP packaging creates Markdown, HTML, manifest, and release zip.