qc

The qc (Quality Gate) skill is an automated code review orchestration tool designed for software engineering teams to maintain high code standards. By spawning five distinct AI agents in parallel, it performs a comprehensive audit of any git diff against the main branch. This multi-layered approach ensures that changes are not only functional but also aligned with existing architectural patterns and maintainability requirements.

• Parallel Agent Architecture: Runs five specialized agents simultaneously covering Code Review, Simplification, Consistency, Robustness, and Scope.

• Automated Standards Enforcement: Validates code naming, error handling, dead code removal, and adherence to DRY principles.

• Context-Aware Reviewing: Leverages existing codebase patterns and file-level context to provide relevant feedback that matches team-specific coding styles.

• Regression Prevention: Evaluates edge cases, API contract stability, and potential side effects to mitigate risks before commits are finalized.

• Workflow Integration: Provides clear results in a table format with pass/fail status, actionable issue lists, and support for rapid iteration loops.

• Input Requirements: The tool expects a git diff output and a task description to provide targeted feedback.

• Best Practices: For large diffs exceeding 500 lines, users are advised to execute the tool file-by-file for higher precision.

• Error Handling: If an agent returns invalid JSON, the tool performs a single automatic retry before escalating a failure.

• Escalation Policy: If a specific issue persists after three attempts at correction, the system automatically escalates the concern to human intervention.

• Implementation: It is built for use in environments leveraging git-based workflows, ensuring that all code reviewed matches project logic, is composable, and isolated where necessary.