Code Review Assistant (Python)

Quick start

• Read the full code to understand intent, architecture, and interactions before commenting.

• Check style, correctness, edge cases, efficiency, security, and maintainability.

• Provide concise, actionable feedback grouped into strengths, issues/risks, and suggestions.

• Include inline code snippets when helpful; ensure recommendations explain the why.

• Follow references/code_review_rules.md for the review checklist and output format.

Workflow

Understand

• Identify the module’s purpose, main flows, and dependencies.

• Note assumptions, data shapes, and external interactions.

Assess quality and correctness

• Validate logic against intent; check edge cases, error handling, and state changes.

• Verify PEP 8 compliance, naming, imports, and type hints.

• Look for dead code, mutable defaults, and duplicated logic.

Evaluate design and efficiency

• Check function/class responsibilities, cohesion, and complexity.

• Consider algorithmic complexity and data structure choices; flag inefficiencies.

• Ensure configuration/secrets are not hardcoded and that dependencies are isolated.

Security and safety

• Identify risky patterns: unsafe I/O, injection risks, use of eval/exec/pickle , or leaked secrets.

• Recommend safer alternatives when applicable.

Testing and output

• Check test coverage, mocking of I/O, and CI hooks if visible.

• Produce the review using the output format in the rules doc with grouped strengths, risks, and suggestions.

Reference

• references/code_review_rules.md : detailed checklist and response format for Python code reviews.