process-code-review

What problem does it solve?

This Skill turns incoming pull request code review feedback into concrete, test-backed code changes while ensuring the resolved items are posted back to the PR in a consistent, rule-compliant way.

Core Features & Use Cases

• Automated review-driven patching: Finds the latest PR(s) for a given task, switches to the PR branch, and maps each review finding to specific code or test changes.
• Reproducer-based test creation: Extracts Faulty Example, Expected Behavior, Test Hint, and Suggested Fix from the originating CR skills and uses them to add failing tests before fixing.
• Iterative convergence loop: Runs the appropriate review skill in-memory until Critical and Moderate findings drop to zero (or max iterations is reached), then publishes only on the final converged run.
• Rule- and language-safety: Applies the required rule sets, preserves canonical English for CR reply comments where specified, and avoids expanding scope beyond review feedback.

Quick Start

Use process-code-review when you need an AI-driven loop that loads the review context for a task’s PR(s), applies only the requested review fixes with diff-scoped testing, and posts resolved items after convergence.