lets-code-review

Automated code review generates structured feedback and confidence-gated findings for code changes, ensuring quality and adherence to project standards before a pull request. Run the review command on your current branch to receive actionable results.

Headless code review mode supports programmatic invocation without interactive prompts. It dynamically selects reviewer personas and outputs confidence-gated findings through a merge and deduplication pipeline for automated processing.

Confidence-gated findings ensure code review results are accurate and actionable by filtering outputs through a merge and dedup pipeline. This mechanism dynamically selects reviewer personas to validate code changes before presenting structured feedback.

Automated code review requires Claude Code or compatible platforms to function. It leverages the environment to dynamically select reviewer personas and process structured feedback across interactive, autofix, report-only, and headless modes.

The best way to review a specific pull request automatically is to provide the PR link as an argument to the code review command. This triggers structured feedback generation and confidence-gated findings for the targeted PR changes.

Autofix mode applies automated corrections to code changes, while report-only mode generates structured feedback without modifying files. Both modes utilize confidence-gated findings and reviewer personas to ensure code quality meets project standards.