Build a standalone, configuration-driven dev supervisor and use the blog repository as the first client.
- Replace default boilerplate with project-specific docs.
- Define architecture boundaries between engine, config, and hooks.
- Capture work in
bd.
- Load a config file from a target repository.
- Start and supervise named processes.
- Watch path groups and classify changes.
- Execute ordered workflow steps.
- Capture session state in a stable file.
- Add an example config for the blog repository.
- Add repo-local hooks for current post and public URL reporting.
- Verify that config can be parsed and the engine can start against the repo.
- Tighten error handling and logging.
- Add health checks and restart sequencing.
- Decide which parts should remain hooks and which should become core features.