Claude:
Review this PR thoroughly (Compare my branch to master branch). For each file changed:
**Code Quality**
- Flag dead code, unreachable branches, redundant logic
- Identify duplicated code that should be extracted
- Spot over-engineering or unnecessary abstractions
- Check for missing null/error handling
**Correctness**
- Logic bugs or off-by-one errors
- Race conditions or async/await misuse
- Edge cases not handled
- Type mismatches or unsafe casts
**Refactoring**
- Rename anything misleading (variables, functions, files)
- Extract functions that are too long or do too many things
- Simplify complex conditionals
- Remove comments that just restate the code
**Tests**
- Missing unit tests for new logic
- Missing edge case coverage
- Tests that test implementation not behavior (brittle)
- Suggest what to test if nothing exists
**Security & Performance**
- N+1 queries or unnecessary re-renders
- Unvalidated inputs or exposed sensitive data
- Memory leaks or missing cleanups
**Output format:**
For each issue, respond as:
- FILE: `path/to/file.ts`
- LINE: (if known)
- SEVERITY: [critical | warning | suggestion]
- ISSUE: what's wrong
- FIX: concrete fix or code snippet
At the end, give a summary verdict: Approve / Request Changes / Needs Discussion.
ChatGPT
Act as a senior software engineer conducting a thorough PR review. (Compare my branch to master branch)
Review the following code changes with a strong focus on correctness, maintainability, and production readiness.
Your review must include:
1. Summary
Briefly explain what the code is doing
Highlight overall quality and any major concerns
2. Critical Issues (if any)
Bugs or incorrect logic
Edge cases not handled
Potential regressions
3. Code Quality & Readability
Naming issues
Unclear or complex logic
Suggestions for simplification or refactoring
4. Architecture & Design
Evaluate structure and modularity
Identify tight coupling or poor abstractions
5. Redundant / Dead Code
Point out unreachable, duplicated, or unnecessary code
Suggest removal or consolidation
6. Performance Considerations
Highlight inefficiencies
Suggest optimizations where meaningful
7. Testing
Evaluate existing tests
Suggest missing test cases (edge cases, failure cases)
Add example unit tests if needed
8. Security & Reliability
Input validation issues
Error handling gaps
Any unsafe patterns
9. Suggested Refactor
Provide improved version(s) of the code where useful
10. Final Verdict
Approve / Request Changes
Clear reasoning
Be specific, actionable, and concise. Prefer concrete code suggestions over vague advice.
Index - LLM Prompts
software