What triggers the Code Review Gate automatically?

The gate triggers automatically for changes in security-sensitive paths (such as auth or api directories), files with more than 500 lines, or functions exceeding a cyclomatic complexity of 15.

Can this skill be bypassed for production emergencies?

Yes, an emergency bypass command is available but requires a mandatory reason and tracking ticket. All bypasses are logged to a permanent audit trail for post-deployment review.

What happens if a code review is rejected?

The skill halts the workflow and returns the project to the implementation phase. You must address the human reviewer's feedback before the gate can be cleared for integration testing.

Can I manually request a code review?

Yes, you can activate the review process manually at any time using the /code-review command or by asking Claude to 'check my implementation'.

Where does the skill store review reports?

Detailed review reports are stored in .claude/reports/code-review-report.json, including timestamps, reviewer comments, and the status of the security checklist.

Code Review Gate

Name: Code Review Gate
Author: turbobeest

byturbobeest

0•

Security & Testing

Enforces human-in-the-loop validation for security-sensitive paths and high-complexity code changes within the Claude Code environment.

The Code Review Gate skill acts as a critical quality control layer between code implementation and integration testing. It automatically triggers for high-risk changes—such as security-sensitive authentication logic, high cyclomatic complexity, or large-scale modifications—ensuring that human oversight is applied where AI automation might miss nuance. By providing a structured review checklist covering security, architecture, and test coverage, this skill helps teams maintain high standards and prevent technical debt or security vulnerabilities from reaching production.

Key Features

01Automatic triggers for security-sensitive file paths like auth, API, and crypto utilities

02Complexity-based gating for functions exceeding cyclomatic complexity thresholds

03Interactive review workflow with commands for approval, rejection, and inline commenting

040 GitHub stars

05Structured review checklists covering security, code quality, and architecture

06Emergency bypass logging with mandatory reason and ticket tracking for audit trails

Use Cases

01Reviewing large-scale refactors that modify core system components or API contracts

02Validating critical security patches or authentication logic before integration

03Ensuring architectural boundaries and layer dependencies are respected in complex PRs

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add turbobeest/squawk code-review-gate

For use in Claude.ai and ChatGPT

Download Skill