Code Review Gatekeeper

About

The requesting-code-review skill implements a disciplined quality assurance layer within the Claude Code environment by requiring a structured review for all substantive code changes. It prevents technical debt and production regressions by dispatching a dedicated review subagent that evaluates code across six critical dimensions—including architecture, security, and testing. This skill ensures that no task is marked as complete or merged until it has passed a rigorous automated verification process, maintaining high standards even in rapid development cycles.

Key Features

• Mandatory review gating for all non-trivial code changes
• Integration with subagent-driven development (SDD) workflows
• Subagent-driven automated code audits across six quality phases
• Structured review status management (Approved, Needs Revision, Blocked)
• Comprehensive audit trails via mandatory usage announcements
• 3 GitHub stars

Use Cases

• Verifying feature implementations against original project plans and specifications
• Validating bug fixes and refactors before merging into the main branch
• Enforcing quality standards during autonomous task execution in agentic workflows