An AI-powered code review assistant that combines local static analysis with deep LLM reasoning to identify bugs, security flaws, and performance bottlenecks.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install code-review-sr
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install code-review-sr using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is CodeReview?
CodeReview is a sophisticated analysis tool designed to bridge the gap between instant linting and deep human-level architectural review. By integrating into your development workflow, this tool provides a dual-layered approach to code quality. It leverages the Openclaw Skills framework to offer developers a seamless way to audit their repositories for vulnerabilities, logical errors, and stylistic inconsistencies without manual overhead.
Whether you are working locally with Ollama or utilizing cloud-based providers like Anthropic and OpenAI, CodeReview ensures that your code is scrutinized through both pattern-matching and contextual reasoning. Its ability to run without external dependencies makes it an ideal choice for teams prioritizing performance and security in their CI/CD pipelines or local IDE environments.
CodeReview Use Cases
- Auditing new pull requests for security vulnerabilities like SQL injection or hardcoded secrets.
- Improving legacy codebases by identifying deeply nested logic and redundant conditions.
- Benchmarking code quality across large directories using the automated scoring system.
- Ensuring consistent coding standards and performance best practices across polyglot projects.
- Providing an offline-first safety net with local regex analysis when AI APIs are unavailable.
How CodeReview Works
- The skill initiates a local pre-pass using regex-based pattern matching to instantly identify high-risk syntax and common security anti-patterns.
- Source code and local findings are aggregated and prepared for the AI engine, with intelligent file truncation to maintain token efficiency.
- The tool communicates with your chosen provider (Anthropic, OpenAI, or Ollama) to perform a deep-reasoning analysis of logic and architecture.
- A comprehensive JSON report is generated, featuring a quality score, categorized issues, and actionable improvement suggestions.
- If the AI service is unreachable, the system gracefully degrades to provide the results of the local static analysis alone.
CodeReview Setup
To integrate this into your project, ensure you have Node.js installed. Set your environment variables for your preferred AI provider to get the most out of Openclaw Skills:
export ANTHROPIC_API_KEY="your-api-key"
# Or for OpenAI
export OPENAI_API_KEY="your-api-key"
Initialize the reviewer in your script:
const { CodeReview } = require('./src/code-review');
const reviewer = new CodeReview({ model: 'anthropic/claude-haiku-4-5' });
const result = await reviewer.review('./src/file.js');
CodeReview Data Schema & Taxonomy
The skill produces a structured JSON output for easy integration into other Openclaw Skills or reporting tools:
| Field | Description |
|---|---|
score |
A 1-10 quality metric based on findings. |
issues |
Array containing severity, line, type, and message. |
suggestions |
List of actionable steps to fix identified problems. |
summary |
A high-level overview of the code's health. |
aiPowered |
Boolean indicating if LLM analysis was successful. |
totalIssues |
Numeric count of all identified problems. |
CodeReview Advanced Features
- Multi-model support including Anthropic, OpenAI, and local Ollama instances for private, air-gapped analysis.
- Concurrent directory processing with configurable batch sizes to handle large-scale repositories efficiently.
- Zero-dependency architecture ensuring the skill is lightweight, secure, and easy to audit.
- Automatic file truncation at 8,000 characters to prevent token limit errors during deep reviews.
- Extensible pattern matching for local analysis that works across multiple programming languages including Python, Go, Rust, and TypeScript.
SKILL.md
Loading
Related Openclaw Skills
Weather and Pollen
A specialized skill for retrieving real-time weather conditions and localized pollen counts using free, no-key APIs.
Vercel
A comprehensive CLI integration for deploying applications, managing projects, and accessing Vercel documentation.
Typefully Social Media Manager
Automate your multi-platform social media presence with advanced scheduling and thread management via the Typefully API.
skills-search
A command-line tool designed to search and discover agent capabilities within the skills.sh registry and ecosystem.
EmailDigest: Daily Email Summary for Agents
EmailDigest is a local-first email categorizer that transforms raw email data into a prioritized daily brief for AI agents.
PixelClaws
A collaborative 1024x1024 pixel art canvas specifically designed for AI agents to coordinate and create art through distributed assignments using Openclaw Skills.
Featured*
