General Analysis Launches MCP Guard

We are excited to launch MCP Guard, the first runtime firewall designed to secure every MCP (Model Context Protocol) tool call against prompt injection attacks.

MCP Guard detects and blocks prompt injection attacks against MCP clients (Cursor, Claude Code, etc.). Read more about the attack here. (Hacker News, Simon Willison Blog, The Prime Time).

Once configured, the MCP Guard acts as a proxy server that sends each tool output to the GA guardrail server to screen for prompt injection. Visit docs.generalanalysis.com to get started.

AI security is a tough problem to tackle, and we are just getting started. For business inquires and collaboration, contact us at info@generalanalysis.com or through our discord server.

Github (ga cli): https://github.com/General-Analysis/ga-cli

Github (proxy MCP server): https://github.com/General-Analysis/mcp-guard

Join the discussion: https://discord.gg/BSsrzPbvyN