auto-review-fix

Automated code review and fixes across a pull request are handled by spawning parallel review agents for changed files, validating findings, and applying fixes iteratively until the codebase is clean or maximum iterations are reached.

An autonomous parallel code review loop is a workflow that spawns specialized agents to review changed files simultaneously, then iteratively validates and applies fixes via dedicated subagents while tracking skipped issues and recording iteration state for audit.

Trigger the autonomous review Skill on a pull request to automatically start a 4-iteration cycle that reviews changed files in parallel, validates findings, and applies fixes until the codebase is clean or the maximum iterations are reached.

The automated code review workflow enforces a structured, crash-recoverable workflow that uses Task agents for reviews and fixes, tracks skipped issues, and records iteration state for audit.

Autonomous fixes are applied to multiple changed files simultaneously by spawning parallel review agents across categories and generating fixes via dedicated subagents while maintaining crash-recovery state.

When the maximum number of iterations is reached, the iterative review-validate-fix cycle stops, leaving the codebase in its current state with skipped issues tracked and iteration state recorded for audit.