review-fix

Automated code review and fixing is handled by spawning multiple specialized agents in parallel to identify issues, automatically resolving quick-fix items, and iterating until code quality is achieved. You simply run the process to clean up minor issues and fix common bugs before submitting your code.

Multi-agent code review works by spawning eight specialized reviewers in parallel to comprehensively analyze your codebase. This parallel execution identifies quick-fix issues for automated resolution and accumulates complex strategic items for user decision.

Yes, automated code fixing supports configurable parameters for iteration count, commit behavior, and base comparison. You can set the maximum iterations to control how many times the review and fix phases loop before stopping.

Complex architectural issues identified during automated code review are accumulated as strategic items and presented to the user for decision. This ensures that quick-fix items are resolved automatically while larger refactoring suggestions require manual approval.

Yes, automated code fixing runs without external dependencies. The process relies entirely on its internal specialized agents to handle review and fix loops, requiring no additional packages or environment setup to identify and resolve code quality issues.

An iterative code review loop is best used after a refactoring session to automatically clean up minor issues, fix common bugs, and identify larger architectural suggestions. It saves significant time by handling repetitive quick-fixes that would otherwise require manual effort.