prompt-defense-audit

What it does

prompt-defense-audit is an MCP (Model Context Protocol) server that integrates MISP (Malware Information Sharing Platform) with Claude and other AI tools. It provides secure access to threat intelligence data through 8 read-only tools while protecting against prompt injection attacks via the promptdefenseaudit library.

Key features

• MISP integration for malware and threat intelligence sharing
• Built-in prompt injection defense mechanisms
• 8 read-only threat intelligence tools: events, attributes, search, tags, feeds, galaxies
• MCP server compatible with Claude, ChatGPT, and AI Studio
• Secure scanning and validation of prompts

How to set up

Install from GitHub repository ppcvote/mispmcpserver. Configure as an MCP server connection in your AI platform settings. Requires MISP instance access credentials. Follow the repository documentation for platform-specific setup (Claude, ChatGPT, AI Studio, or Antigravity).