reversa-docs

Generate offline HTML documentation by running the pipeline to transform extracted architecture and feature specs into a self-contained mini-site. It bundles local vendor assets and writes navigable pages strictly under .reversa/documentation/.

Yes, generated project documentation works offline without remote CDN dependencies. The process bundles required vendor libraries into local assets/vendor/ directories, ensuring file:// safe viewing for all architecture and glossary pages.

When source files are missing during static documentation mini-site generation, the pipeline gracefully detects absent inputs and omits dependent pages. It produces a minimal but usable site instead of failing the entire build process.

Regenerate specific documentation pages without rebuilding the entire architecture mini-site using per-page or per-agent regeneration. The system supports targeted reruns and creates backups for destructive writes to preserve previous state.

Generating offline documentation requires reading .reversa/state.json and .reversa/documentation/.config.json if present. It also detects expected sources from references/expected_sources.yaml to orchestrate the documentation pipeline correctly.

The best way to publish extracted knowledge as a navigable project documentation bundle is orchestrating specialized agents in sequence. The pipeline runs Mapper, Analyst, Storyteller, and Publisher agents to produce consistent architecture, metrics, and glossary outputs.