MCP Code Review Server
AI code review in your editor—no SaaS fees, no uploads
Start new thread

MCP Code Review Server - AI code review in your editor—no SaaS fees, no uploads

by•
MCP Code Review Server brings professional-grade code review to your AI coding assistant. Connect it to Claude Code, Cursor, Cline, or any MCP client. Features include OWASP security scanning, bug and vulnerability detection, N+1 query analysis, and performance profiling — all with structured severity ratings and actionable fix suggestions. Runs entirely locally. One-time $49 purchase, lifetime updates.

Add a comment

Replies

Best
Maker
📌
I built MCP Code Review Server because I was tired of uploading code to third-party SaaS tools just to get a security review. This MCP server integrates directly with Claude Code, Cursor, and Cline — so you can ask your AI to review code right in your editor. It runs entirely locally with OWASP security scanning, bug detection, and performance analysis. Key features: • OWASP Top 10 scanning built-in • Structured reviews with severity ratings • Git diff review before merging • Works with any MCP client • Runs locally — no code leaves your machine • One-time purchase, lifetime updates GitHub: Would love to hear your feedback!

Love that it runs locally and the OWASP scanning is included out of the box. One thing that would really level it up for my team: a way to define custom rule sets per repo, so we can enforce our own style and architecture conventions alongside the built-in checks. Right now the defaults are solid but every codebase has its own quirks that generic rules miss.

Maker

 Thanks for the great suggestion! Custom rule sets per repo is exactly the kind of thing we want to build next. The OWASP and generic checks are the foundation, but letting teams layer on their own conventions would make it a proper code review companion. I've added this to our roadmap — will follow up when there's something to test. Appreciate you taking the time!

Runs locally is honestly a huge plus for me. One thing that would make it even better is adding team-shared rule profiles, basically let my whole squad sync their custom review rules and severity thresholds from one config so we stop getting inconsistent feedback across different machines.

Maker

 Team-shared rule profiles — yes, this keeps coming up and it makes total sense. A single config that syncs across the team would solve that inconsistent feedback problem nicely. Between this and the custom rules suggestion earlier, it's clear teams want the safety net of built-in checks + the flexibility of their own conventions. Definitely on the roadmap. Thanks for reinforcing it!