domains / software-engineering / code-review / code-reviewer

code-reviewer

Skill card

combined

The four copies are not reproductions of one procedure. Jeffallan/claude-skills defines a manual, structured review workflow with a specific output template and constraints. foryourhealth111-pixel/Vibe-Skills defines a routing boundary and relies on automated Python scripts (pr_analyzer, code_quality_checker, review_report_generator). zhnnky329/MathModeling-skills is a narrow router for Python vs MATLAB modeling code that outputs a JSON review contract. rmyndharis/antigravity-skills is a broad, tool-centric expert persona covering AI tools, static analysis, and infrastructure. Because they disagree on the primary mechanism (manual workflow vs scripts vs language router vs tool persona) and each contains material steps the others lack, crowning one would hide real alternatives.

Repositories

Other copies

Combined skill

# Code Reviewer

## Routing
1. Determine the review scope:
   - If the code is Python or MATLAB modeling code from `qx_code_plan.md`, route to the language-specific reviewer (Python or MATLAB) and use the JSON review contract.
   - If the task is a dedicated OWASP/security audit, route to `security-reviewer`.
   - If the task is handling existing review comments, route to `receiving-code-review`.
   - Otherwise, proceed with the general code review workflow below.

## General Review Workflow
1. **Context**: Read the PR description and summarize the intent in one sentence. If unclear, ask the author.
2. **Structure**: Review architecture and design decisions. Check for adherence to existing patterns.
3. **Details**: Check for bugs, security vulnerabilities (SQL injection, XSS, etc.), N+1 queries, and performance issues.
4. **Tests**: Validate test coverage and quality. Ensure edge cases are covered.
5. **Feedback**: Produce a structured report.

## Automated Analysis (Optional)
If available, run the following scripts to assist the review:
- `python scripts/pr_analyzer.py <project-path>`
- `python scripts/code_quality_checker.py <target-path>`
- `python scripts/review_report_generator.py [arguments]`

## Output Format
### General Review Report
1. **Summary**: One-sentence intent recap + overall assessment.
2. **Critical Issues**: Must fix before merge (bugs, security, data loss).
3. **Major Issues**: Should fix (performance, design, maintainability).
4. **Minor Issues**: Nice to have (naming, readability).
5. **Positive Feedback**: Specific patterns done well.
6. **Questions for Author**: Clarifications needed.
7. **Verdict**: Approve / Request Changes / Comment.

### Modeling Code Review (JSON)
If routing to modeling code, output JSON with these named checks:
- `syntax`
- `input_contract`
- `method_alignment`
- `reproducibility`
- `output_contract`

## Constraints
- Provide specific, actionable feedback with code examples.
- Prioritize feedback (critical → minor).
- Do not nitpick style when linters are configured.
- Acknowledge author reasoning for non-obvious choices.
- Do not infer success because scripts exist; validate outcomes.