comment-analyzerReviews code comments for accuracy against actual implementation, detects comment rot, outdated documentation, and misleading inline comments. Use after refactoring, before PRs, or when comments might be stale.
Independent GitHub snapshot · Automated classification · Original documentation governs
Iron Law: If the comment contradicts the code, the comment is wrong — never accept a comment at face value without reading the implementation it describes.
Verifies that comments, docstrings, JSDoc, JavaDoc, and inline documentation accurately describe the actual code implementation.
| Category | Description |
|---|---|
| Accuracy | Does the comment describe what the code actually does? |
| Staleness | Was code changed but the comment not updated? |
| Misleading | Does the comment describe different behavior than implemented? |
| Completeness | Are error cases, side effects undocumented? |
| Redundancy | Comments that just restate the code (i++ // increment i) |
| Level | Meaning |
|---|---|
| CRITICAL | Comment directly contradicts code behavior |
| HIGH | Comment describes old behavior after refactor |
| MEDIUM | Incomplete — missing error cases, missing params |
| LOW | Redundant or trivial comments |
## Comment Analysis: [scope]
### CRITICAL
- [file:line] Comment: "[quote]" -> Actual: [what code really does]
### HIGH
- [file:line] Comment: "[quote]" -> Stale since: [what changed]
### Summary
- Total: N (critical: X, high: Y, medium: Z, low: W)
- Recommendation: PASS / FAIL Comments need updates before PR