A curated list of AI tools and resources for developers, see the AI Resources .

MCP Scanner

A tool to scan MCP servers and tools for potential security issues, using multi-engine analysis and customizable reporting.

Introduction

MCP Scanner is a security scanning toolkit from Cisco AI Defense that inspects MCP (Model Context Protocol) servers, tools, prompts, and resources for potential vulnerabilities. It combines multiple analyzers—YARA rules, LLM-as-judge, and Cisco AI Defense analyzers—into flexible pipelines that can run independently or together.

Key features

  • Multi-engine analysis: combine YARA, LLM judgment, and Cisco AI Defense engines.
  • Multiple modes: CLI, REST API server, and SDK for integration and automation.
  • Customizable rules and reporting formats for auditability and workflow integration.

Use cases

  • Security audits of public or internal MCP services.
  • Integrating scanning into CI/CD pipelines to catch vulnerabilities early.
  • Incident investigation and baseline security checks for security teams.

Technical highlights

  • Python-based implementation with async REST API, pluggable analyzers, and benchmark tooling.
  • Supports OAuth and Bearer Token authentication, and can integrate cloud LLM providers to enhance detection.
  • Comprehensive docs and benchmark suite for reproducible testing and performance evaluation.

Comments

MCP Scanner
Resource Info
🌱 Open Source 🛠️ Dev Tools 🛡️ Safety