reverse-engineering-specs

Documenting legacy code behavior for safe refactoring involves exhaustively tracing code paths, data flows, and state mutations to generate implementation-free behavioral specifications. This captures actual functionality, including bugs, enabling developers to understand and modify brownfield projects safely.

Implementation-free behavioral specification generation is the process of creating human-readable documentation that describes what code does rather than how it works. It traces every code path and data flow to document actual behavior exhaustively without referencing internal implementation details.

Yes, capturing existing bugs when reverse engineering code specifications is supported by documenting actual behavior, including defects, in a separate KNOWN_ISSUES.md file. This ensures known issues are recorded exhaustively alongside behavioral specs for safe brownfield project modifications.

The best way to generate documentation for an undocumented microservice is using parallel subagents to comprehensively investigate code paths and state mutations. This approach produces exhaustive behavioral specifications that allow new developers to understand functionality without deciphering complex code.

Yes, reverse engineering specifications works for brownfield project feature additions by exhaustively tracing existing code paths and data flows to generate behavioral specs. This allows developers to understand current functionality and add features safely without breaking existing behavior.

Yes, parallel subagents are required for comprehensive code investigation when tracing code paths and data flows. They enable exhaustive analysis of existing codebases to document actual behavior and state mutations accurately for behavioral specification generation.