reversa-detective

To extract business rules from a legacy codebase, this skill analyzes complex conditionals, validations, enums, and comments to surface implied domain constraints. It generates a domain.md file mapping these buried rules into clear, operational specifications.

Reverse engineering retroactive ADRs from Git history involves analyzing commits, fixes, refactors, and reverts to reconstruct the real decision rationale. This skill outputs an adrs/ directory containing evidence-backed architectural decision records with confidence labeling.

Yes, you can generate state machines and RBAC permission matrices by analyzing observed statuses and access models within legacy code. This skill produces state-machines.md and permissions.md files documenting the operational constraints and entity structures.

Documenting operational constraints for AI agents requires transforming implicit buried knowledge into safe specifications. This skill reads the Reversa state.json output_folder and doc_level to produce traceable artifacts like domain.md and permissions.md with evidence-backed confidence labeling.

Reversa-detective requires no external dependencies or components to run. It functions by reading the Reversa state.json configuration to identify the output folder and documentation level before generating the interpreted domain artifacts.

A limitation of reconstructing domain logic from inferred entity states is that confidence levels vary based on available evidence. The skill addresses this by applying evidence-backed confidence labeling to all generated outputs, ensuring traceability for AI agent specifications.