Code Review Helper

Problem

Code review often gets stuck on two costs: manually reading diffs is slow, while generic linters only catch surface-level style issues. Code Review Helper targets PRs or local changes in multilingual projects, focusing review on code standards, security vulnerabilities, performance issues, and potential bugs so engineers can decide more quickly whether a change is ready to merge.

How It Works

The skill analyzes code changes and supports Python, JavaScript, TypeScript, Go, Rust, and Java. It can work with Git Diff input and evaluate issues with surrounding context instead of returning generic advice. Typical output includes:
- Style issues: naming, structure, readability, and consistency risks;
- Security and performance: potential vulnerabilities, resource usage, and inefficient implementations;
- Optimization suggestions: actionable fixes and example code where useful.
It also performs complexity analysis to help determine whether functions, control flow, or dependencies are growing too large.

Boundaries

It is best treated as a review aid, not a compiler, full static security scanner, or CI gate. Complex business logic, framework conventions, and internal standards still require repository context and human confirmation. For large diffs, missing context, or specialized language ecosystems, split the review by module and avoid treating one-pass suggestions as the final verdict.