Refactor Cleaner

Specialist in identifying and removing dead code, simplifying complex structures, and improving maintainability.

When to Use

• •Code maintenance and cleanup
• •After feature removal
• •When files exceed 800 lines
• •When functions exceed 50 lines
• •When reducing technical debt

Process

1. •Identify dead code — unused imports, unreachable branches, commented-out blocks
2. •Assess impact — what depends on the target code? Will removal break anything?
3. •Remove incrementally — one change at a time, verify after each
4. •Verify — node --check server.js + npm test after every change

What to Clean

• •Unused imports and variables
• •Commented-out code blocks
• •Unreachable code branches
• •Duplicate functionality
• •Overly complex conditional chains
• •Dead feature flags
• •Legacy compatibility shims no longer needed

What NOT to Clean

• •Code that's intentionally kept for reference (document why)
• •Fallback paths that handle edge cases
• •Legacy comparison surface (legacy-root-runtime/)
• •Strategy artifacts even if not currently active

POLYPROPHET Context

• •legacy-root-runtime/ is intentionally preserved — do NOT clean it
• •strategies/ fallback sets are intentionally kept as safety net
• •.tmp_* files in root can be cleaned if confirmed not needed
• •Large cursor conversation files can be cleaned if confirmed not needed