code-comments

Comments Extraction Skill

Extract comment locations from git-changed files for analysis by the comment-cleaner agent.

What This Skill Does

• Detects files changed in a git scope (branch or unstaged)
• Extracts all comments with line numbers and context
• Outputs structured JSON for downstream processing
• Supports multiple programming languages

When to Use

Activate this skill when:

• Preparing for comment cleanup operations
• Auditing code documentation coverage
• Analyzing comment patterns in a codebase
• Working with the comment-cleaner agent

Supported Languages

Usage

Extract Comments from Git Scope

# Default: files changed since branch diverged from main
rp1 agent-tools comment-extract branch main

# Only unstaged files (pre-commit use case)
rp1 agent-tools comment-extract unstaged main

# Extract from commit range with line-scoped filtering
rp1 agent-tools comment-extract "abc123..def456" main --line-scoped

Output Format

{
  "success": true,
  "tool": "comment-extract",
  "data": {
    "scope": "branch",
    "base": "main",
    "filesScanned": 12,
    "linesAdded": 150,
    "comments": [
      {
        "file": "src/auth.py",
        "line": 45,
        "type": "single",
        "content": "# Check if user is active",
        "contextBefore": "def validate_user(user):",
        "contextAfter": "    if user.is_active:"
      }
    ]
  }
}

Output Fields

Error Handling

The tool handles:

• Git command failures (not a repo, invalid branch)
• Missing files (deleted in working tree)
• Binary files (automatically skipped)
• Encoding issues (UTF-8)

Error output format:

{
  "success": false,
  "tool": "comment-extract",
  "data": {
    "scope": "branch",
    "base": "main",
    "filesScanned": 0,
    "linesAdded": 0,
    "comments": []
  },
  "errors": [{ "message": "Not a git repository" }]
}

Integration

This skill is used by the comment-cleaner agent to:

1. Get a manifest of all comments in scope
2. Avoid reading entire files for comment detection
3. Process comments efficiently with context

Limitations

• Does not detect comments inside string literals (best effort)
• Multi-line string docstrings in Python are not extracted (intentional - docstrings are kept)
• Very large files (>10000 lines) are skipped to prevent memory issues