Documentation consistency reviewer that checks alignment between code implementation and documentation...
Systematically identify all "outdated" or "inconsistent with implementation" descriptions in README + docs/, outputting ≥30 issue items.
# Scan scope
- README.md (root directory)
- docs/**/*.md (all documentation)
- Contract files: OpenAPI/proto/GraphQL schema/TS types
For each document:
See checklist.md for detailed review checklist.
| Level | Definition | Example |
|---|---|---|
| P0 | Security issue/serious misleading | Docs say sandbox enabled but code doesn't enable it |
| P1 | Core functionality inconsistency | Following docs leads to failure |
| P2 | Incomplete examples/naming inconsistency | Doesn't directly block usage |
| P3 | Wording/formatting/link minor issues | Doesn't affect functionality |
| Pending Evidence | Suspicious but insufficient evidence | Needs further investigation |
See output-format.md for detailed templates.
### [Title]
- **Severity**: P0/P1/P2/P3/Pending Evidence
- **Location**: `<file_path>:<line_number>`
- **Evidence**:
- Documentation: [quote]
- Code: [quote]
- **Impact**: [Misleading consequences]
- **Suggestion**: [Minimal fix]
- **Related Principle**: Code is truth/Contracts first/Security default tightening/...
## Review Conclusion
- **Verdict**: Pass/Conditional Pass/Fail
- **Summary**: P0:x P1:x P2:x P3:x Pending:x
- **Fix Priority**: P0 → P1 → P2 → P3
For acceleration, split by following dimensions for parallel multi-agent execution:
Deduplication and unified severity rating needed when aggregating.
After review completion, output doc-consistency.md report file, and also output doc-consistency.json (structured issue list for aggregation/deduplication/statistics).