Detailed user documentation is split by topic so the landing page can stay short without losing the durable product manual.
For quick command lookup, start with the cheatsheet.
| Topic | Document |
|---|---|
| Install paths and quickstart modes | docs/install.md |
| Spec format, workspace config, sources, selectors, artifacts | docs/configuration.md |
| Runtime setup, indexing, output formats, MCP server | docs/runtime.md |
| Compliance, reports, CI, command reference | docs/workflows.md |
| Agent usage guidance and reusable skill package | skills/pituitary-cli/SKILL.md, repo-context.md, AGENTS.md |
- New user evaluating Pituitary on an existing repo:
- docs/install.md
- docs/cheatsheet.md
- Run
pituitary explain-file PATHfirst when scope or source matching looks wrong
- Repo owner wiring Pituitary into CI:
- Contributor changing config or source behavior:
- User tuning local retrieval quality:
Pituitary is specification-first. The core workflow is local filesystem input, deterministic indexing, and CLI-first analysis. Optional runtime-backed retrieval and MCP integration sit on top of that core rather than replacing it.
See: