Code Review Worker Agent

Overview

The Code Review Worker Agent provides a robust, multi-faceted automated system for assessing code quality. It goes beyond simple linting by incorporating weighted scoring across several critical dimensions: security, complexity, style adherence, and test coverage.

The agent uses a defined scoring algorithm (Security 40%, Complexity 30%, Style 20%, Coverage 10%) to provide an actionable pass/fail recommendation, ensuring that submitted code meets predefined quality thresholds before integration into the main codebase.

Capabilities

• Comprehensive Analysis: Performs deep scans covering structural complexity and known security patterns (including OWASP checks).
• Weighted Scoring: Calculates a holistic score based on configurable weights for different quality metrics.
• Automated Feedback: Generates detailed ReviewReport objects, pinpointing specific findings and suggesting necessary remediation steps.
• Iterative Review: Supports up to two re-review iterations, allowing the agent to assess improvements made by developers after initial feedback.

Example Use Cases

1. Pre-Merge Gatekeeping: Integrate this agent into your CI/CD pipeline to automatically block merges if the calculated score falls below the 'Good' threshold (70).
2. Vulnerability Triage: Run it against new feature branches specifically targeting high-risk security findings identified by the OWASP scanner.
3. Code Debt Assessment: Periodically run reviews on older, untouched modules to identify accumulating complexity and style drift.