MCPThreat

andrewbrooks/MCPThreat

★ 0 stars TypeScript AI/LLM Updated today

Threat modeling tool for MCP servers

View on GitHub → 🔍 Audit Wallet Slippage →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "mcpthreat": {
      "command": "npx",
      "args": [
        "-y",
        "andrewbrooks/MCPThreat"
      ]
    }
  }
}

README Excerpt

**MCPThreat is a highly functional prototype for rapid threat modeling of Model Context Protocol (MCP) servers** MCP connects LLM agents to real tools, data, and downstream systems, and in doing so introduces a class of risks that traditional threat modeling doesn't cover well such as prompt injection, tool poisoning, token passthrough, context bleed between tenants, and more. MCPThreat turns reasoning about those risks into a structured, repeatable workflow. You map the trust boundaries, chart

MCPThreat

Quick Install

README Excerpt

Related AI/LLM Servers