Code Review Skill
Review checklist for code in this project.
Review Checklist
Correctness (CRITICAL)
- • Logic is correct for all inputs
- • Return values are correct types
- • Algorithms are implemented correctly
- • No off-by-one errors
Edge Cases
- • Empty inputs handled
- • Null/None/undefined handled
- • Zero values handled (division by zero)
- • Boundary values handled
- • Very large/small values handled
- • inf/NaN values handled (if applicable)
Error Handling
- • Appropriate exceptions/errors raised
- • Error messages are helpful (describe what went wrong)
- • Errors don't leak sensitive information
- • Errors are recoverable where possible
Code Patterns
- • Follows existing patterns in the codebase
- • Consistent naming conventions
- • Appropriate use of abstractions
- • No unnecessary complexity
Code Quality
- • Type hints on public functions (if typed language)
- • Docstrings/comments on public functions
- • No code duplication
- • Reasonable function/method length
Testing
- • Happy path tests exist
- • Edge case tests exist
- • Error case tests exist
- • Tests are meaningful (not just coverage)
Review Output Format
markdown
### File: `path/to/file.py` **Reviewed by**: agent **Status**: APPROVED / ISSUES FOUND / NEEDS CHANGES **Issues Found**: 1. [CRITICAL] Description... 2. [IMPORTANT] Description... 3. [MINOR] Description... **Recommendations**: - Suggestion 1 - Suggestion 2
After Review
Update CONTINUITY.md:
- •Add review findings to CODE REVIEW section (if exists)
- •Mark review task as DONE in TASK QUEUE
- •Add session log entry