brewcode:rules

Project rule syncing works by processing information from KNOWLEDGE.jsonl files, specified file paths, or session learnings to update the `.claude/rules/` directory. It maintains `avoid.md` and `best-practice.md` files, including domain-specific variants, ensuring project guidelines stay consistent.

The best way to document anti-patterns from a code review is using a Skill that extracts them from the current session and automatically adds them to your project's `avoid.md` file. This ensures identified anti-patterns are saved as persistent project guidelines for future reference.

Maintaining best-practice.md files for specific domains is supported by generating domain-specific variants within the `.claude/rules/` directory. The Skill processes knowledge sources to create and update these files, keeping domain-specific guidelines organized and current.

Yes, processing project guidelines requires a `bc-rules-organizer` agent for organizing the information. Additionally, validation scripts are required to perform integrity checks on the generated rule files to ensure correctness.

Yes, you can list existing project rules before adding new best practices. Running the Skill lists all existing rule files in the project, allowing you to review current guidelines before synchronizing new knowledge or session learnings into the `.claude/rules/` directory.