Code Review with LLMs

LLM-Powered Code Review LLMs can review code for bugs, style issues, security vulnerabilities, and best practice violations.

Review Approaches

Comprehensive Review

def review_code(code: str, language: str) -> str:
    return llm.generate(f"""
Review this {language} code for:
1. Bugs and logical errors
2. Security vulnerabilities
3. Performance issues
4. Code style and readability
5. Best practice violations

Code:
```{language}
{code}

Provide specific line numbers and suggested fixes. """)

### Focused Reviews
```python
# Security-focused
def security_review(code: str) -> str:
    return llm.generate(f"""
Analyze for security vulnerabilities:
- SQL injection
- XSS
- Authentication issues
- Secrets in code
- Input validation

Code: {code}
    """)

# Performance-focused
def perf_review(code: str) -> str:
    return llm.generate(f"""
Identify performance issues:
- N+1 queries
- Memory leaks
- Inefficient algorithms
- Unnecessary allocations

Code: {code}
    """)

PR Review Automation

def review_pr(diff: str, context: str) -> dict:
    return llm.generate(f"""
Review this PR diff. Context: {context}

Diff:
{diff}

Return JSON with:
- summary: what the change does
- issues: list of problems found
- suggestions: improvements
- approval: approve/request_changes/comment
    """)

Integration Points

Comparison with Static Analysis

Best Practices

• Use LLM review to supplement, not replace, other tools
• Provide project context (conventions, dependencies)
• Review LLM suggestions before applying
• Fine-tune prompts for your codebase
• Cache reviews for unchanged files