review

Multi-perspective code review evaluates quality, security, performance, architecture, simplicity, and data integrity against project rules. It classifies findings by P1, P2, or P3 severity and provides actionable feedback with specific file and line references.

Yes, code review supports reviewing staged changes, specific files or directories, and PR branches. You can target a feature branch by passing the branch name to analyze changes against project rules and agent-specific guidelines.

Severity classifications in code analysis rank issues by impact: P1 denotes critical issues, P2 indicates important problems, and P3 represents suggestions. This helps prioritize fixes for security vulnerabilities, performance bottlenecks, and architectural flaws.

Automated code analysis works with custom code standards by evaluating code against defined project rules and agent-specific guidelines. It identifies critical issues across multiple perspectives and outputs actionable feedback with precise file and line references.

The best way to ensure quality before merging is running a multi-perspective code review on the PR branch. This identifies critical security, performance, and architecture issues, classifying them by severity to ensure only high-quality code is integrated.

Code review focuses on data integrity and architecture alongside security because comprehensive analysis requires evaluating overall code health. Analyzing simplicity, performance, and quality ensures robust software engineering practices beyond just fixing critical vulnerabilities.