open-code-review

Overview

Open Code Review is an MCP server that provides deep code analysis specifically for AI-generated code. It detects issues that traditional linters miss, including hallucinated imports, stale APIs, over-engineering patterns, and security anti-patterns across 6 programming languages. The tool combines structural analysis, semantic embedding analysis, and LLM deep scanning to provide comprehensive code quality assessment.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• ocr_scan

  Perform a quality gate scan on source code to detect issues like hallucinated imports, stale APIs, and security anti-patterns.

  Try:Scan my TypeScript code for any issues with hallucinated importsTry:Perform a deep scan of the source code with L3 analysisTry:Check the Python files in src/ for security anti-patterns
• ocr_heal

  AI auto-fix for issues detected during scanning, with preview before applying changes.

  Try:Fix the issues found in the scan but show me before applyingTry:Auto-fix the detected security issues with OllamaTry:Apply fixes to the code and set up IDE rules
• ocr_explain

  Get detailed explanations about specific issues detected during code review.

  Try:Explain this hallucinated import issueTry:What does this stale API detection mean?Try:Give me more details about the security anti-pattern found

Comparable tools

Installation

npm install -g @opencodereview/cli

npm install -g @opencodereview/mcp-server

{
  "mcpServers": {
    "open-code-review": {
      "command": "npx",
      "args": ["-y", "@opencodereview/mcp-server"]
    }
  }
}

FAQ

What AI tools does this work with?

Works with all AI coding assistants including GitHub Copilot, Cursor, Claude Code, Codex, Supermaven, Aider, and any tool that generates code.

Can I run this locally?

Yes, it supports local analysis through Ollama or any OpenAI-compatible provider without sending code to cloud services.

Compare open-code-review with