Clean Comments Command

Remove useless comments from files, keeping only meaningful ones.

<removal_criteria> Remove comments that:

• Restate obvious code behavior (the code already says this)
• Are commented-out code blocks (use git history instead)
• Add no information beyond what the code expresses
• Are redundant with function/variable names </removal_criteria>

<preservation_criteria> Keep comments that:

• Explain non-obvious logic or algorithms
• Document magic numbers/constants with reasoning
• Define interfaces, types, or contracts for public APIs
• Provide important context not evident from code structure
• Are copyright headers or license blocks </preservation_criteria>

<workflow> 1. **Identify files:** - Use `$ARGUMENTS` if provided, otherwise `git diff --name-only` - Focus on code files (js, ts, py, go, rs, java, etc.)

2. Read files in parallel:

   • Read multiple files simultaneously when possible
   • NEVER edit a file you haven't read first

3. Evaluate each comment:

   • Consider whether it adds value beyond the code itself
   • Prefer removal when uncertain—removing marginal comments costs little, but keeping noise accumulates technical debt

4. Apply edits:

   • Use Edit tool to remove useless comments only
   • Do NOT add new comments, refactor, or make other changes
   • Preserve formatting and structure

5. Report results: List files processed with comments removed vs kept counts

   </workflow>

<guidelines> - Be aggressive with removal—code should be self-documenting - Never remove JSDoc/docstring API documentation for public interfaces - Focus on inline comments and implementation details - If a file has no useless comments, note it and move on </guidelines>