code-reviewer

What This Skill Does

The code-reviewer skill is a robust, automated framework integrated into OpenClaw designed to streamline the software development lifecycle. By utilizing three core sub-tools—the PR Analyzer, the Code Quality Checker, and the Review Report Generator—it acts as an expert pair programmer. It evaluates code repositories across multiple languages, including TypeScript, JavaScript, Python, Go, Swift, and Kotlin. The tool automates the tedious aspects of code reviews by identifying security vulnerabilities, enforcing clean code architecture (SOLID principles), and quantifying technical debt through cyclomatic complexity and code smell detection. It transforms raw diff data into actionable, high-level reports that prioritize review efforts based on critical risk assessments.

Installation

To integrate this skill into your environment, run the following command within your terminal or OpenClaw interface:

clawhub install openclaw/skills/skills/alirezarezvani/code-reviewer-2

Ensure your environment has the necessary Python runtime dependencies configured to support the execution scripts included in the repository.

Use Cases

• Automated Pull Request Reviews: Use this skill during continuous integration pipelines to perform an immediate, objective sanity check on incoming code before human reviewers spend time on it.
• Technical Debt Assessment: Run the Code Quality Checker against legacy codebases to identify 'God classes' and overly complex functions that require refactoring.
• Security Auditing: Rapidly scan for hardcoded credentials, API keys, and common injection patterns that might be accidentally introduced in a feature branch.
• Onboarding and Mentorship: Generate comprehensive review reports to provide junior developers with immediate, constructive feedback on their coding standards.

Example Prompts

1. "@code-reviewer, analyze the current branch against main and give me a risk assessment for the latest pull request."
2. "Perform a code quality audit on the /src/services directory. Identify any SOLID violations or functions that exceed the 50-line threshold."
3. "Generate a summary review report for the current repository and save it as a markdown file for the team to review."

Tips & Limitations

• Threshold Tuning: While the tool comes with default thresholds (e.g., >50 lines for functions), remember that these are heuristics. High complexity scores do not always mean the code is 'bad,' but rather that it warrants manual inspection.
• Context Awareness: The PR Analyzer works best when comparing two distinct git branches. Always ensure your git history is updated before running the analyzer.
• False Positives: Static analysis tools may occasionally flag complex logic that is intentional. Use the tool to augment human intuition rather than replacing it entirely.
• Language Constraints: Ensure the code being scanned is one of the supported languages; attempting to scan unsupported languages may yield partial or inaccurate results.