Medusa
AI-first security scanner with 76 analyzers, 9,600+ detection rules, and repo poisoning detection for AI/ML, LLM agents, and MCP servers. Scan any GitHub repo with: medusa scan --git user/repo
What is Medusa?
Medusa is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to ai-first security scanner with 76 analyzers, 9,600+ detection rules, and repo poisoning detection for ai/ml, llm agents, and mcp servers. scan any github repo with: medusa scan --git user/repo
AI-first security scanner with 76 analyzers, 9,600+ detection rules, and repo poisoning detection for AI/ML, LLM agents, and MCP servers. Scan any GitHub repo with: medusa scan --git user/repo
This server falls under the Security category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- AI-first security scanner with 76 analyzers, 9,600+ detectio
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx medusaConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Medusa
Medusa is an AI-first security scanner that runs 79 analyzers with 40,000+ detection patterns to find vulnerabilities, leaked credentials, and supply-chain threats in local codebases or remote GitHub repositories. It was built specifically to detect threats relevant to AI development environments, including prompt injection in configuration files, MCP server poisoning, RAG attacks, and misuse of AI editor context files such as .cursorrules. Developers and security teams use it to shift security left, integrate into CI/CD pipelines with SARIF reporting, and proactively harden AI-powered codebases before deployment.
Prerequisites
- Python 3.10 or higher installed
- pip or a virtual environment tool available
- Git installed (for scanning remote repositories)
- An MCP-capable client such as Claude Desktop or Claude Code (for MCP integration)
Install Medusa
Install the medusa-security package from PyPI. Using a virtual environment is recommended to keep dependencies isolated.
python3 -m venv medusa-env
source medusa-env/bin/activate
pip install medusa-securityRun the interactive setup wizard
Run medusa init to configure Medusa for your environment, including IDE integrations and fail thresholds. It will create a .medusa.yml configuration file in your project root.
medusa init
# Or configure all IDE integrations at once
medusa init --ide allScan a local directory or remote repository
Use medusa scan to analyze a local codebase, or pass --git with a GitHub repository path to scan a remote repo directly. Use --quick to scan only changed files for speed in CI.
# Scan the current directory
medusa scan .
# Scan a remote GitHub repository
medusa scan --git owner/repo
# Quick scan (changed files only)
medusa scan . --quick
# Scan for leaked secrets
medusa secrets scanConfigure fail threshold for CI/CD
Set the severity level at which Medusa will exit with a non-zero code, enabling pipeline gating. Options are critical, high, medium, or low.
medusa scan . --fail-on highGenerate a report
Export scan results in HTML, JSON, Markdown, or SARIF format for integration with GitHub Advanced Security, Jira, or your security dashboard.
medusa scan . --format html
medusa scan . --format sarifMedusa Examples
Client configuration
Claude Desktop JSON config to run Medusa as an MCP server using npx.
{
"mcpServers": {
"medusa": {
"command": "npx",
"args": ["medusa"],
"env": {}
}
}
}Prompts to try
Example prompts when Medusa is connected as an MCP server or used from the CLI.
- "Scan my current project for security vulnerabilities and show critical findings"
- "Check the github.com/org/repo repository for supply chain attacks and MCP poisoning"
- "Find any leaked API keys or credentials in this codebase"
- "Scan my AI editor configuration files for prompt injection risks"
- "Run a security scan and generate a SARIF report for GitHub Advanced Security"Troubleshooting Medusa
medusa command not found after pip install
Ensure your virtual environment is activated (source medusa-env/bin/activate) or that the Python scripts directory is in your PATH. On macOS/Linux, try python3 -m medusa as a fallback.
Remote git scan fails or returns no results
Make sure Git is installed and accessible from your PATH. For private repositories, ensure you have authenticated with GitHub (e.g., via gh auth login or SSH keys) before running medusa scan --git.
False positives on minified JS or vendored files
Add exclusions to .medusa.yml under exclude.paths or exclude.files. Common exclusions include node_modules/, venv/, .git/, and patterns like '*.min.js'.
Frequently Asked Questions about Medusa
What is Medusa?
Medusa is a Model Context Protocol (MCP) server that ai-first security scanner with 76 analyzers, 9,600+ detection rules, and repo poisoning detection for ai/ml, llm agents, and mcp servers. scan any github repo with: medusa scan --git user/repo It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Medusa?
Follow the installation instructions on the Medusa GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Medusa?
Medusa works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Medusa free to use?
Yes, Medusa is open source and available under the AGPL-3.0 license. You can use it freely in both personal and commercial projects.
Medusa Alternatives — Similar Security Servers
Looking for alternatives to Medusa? Here are other popular security servers you can use with Claude, Cursor, and VS Code.
Casdoor
★ 13.6kAn open-source Agent-first Identity and Access Management (IAM) /LLM MCP & agent gateway and auth server with web UI supporting OpenClaw, MCP, OAuth, OIDC, SAML, CAS, LDAP, SCIM, WebAuthn, TOTP, MFA, Face ID, Google Workspace, Azure AD
ghidraMCP
★ 9.0kAn Model Context Protocol server that enables LLMs to autonomously reverse engineer applications by exposing Ghidra's decompilation and analysis tools. It allows AI agents to list code structures, rename methods, and analyze binaries directly through
HexStrike AI
★ 8.9kHexStrike AI MCP Agents is an advanced MCP server that lets AI agents (Claude, GPT, Copilot, etc.) autonomously run 150+ cybersecurity tools for automated pentesting, vulnerability discovery, bug bounty automation, and security research. Seamlessly b
IDA Pro MCP
★ 8.7kEnables AI-assisted reverse engineering in IDA Pro by providing tools to analyze binaries, decompile functions, manage comments, search patterns, and interact with the IDA database through natural language.
Anthropic Cybersecurity Skills
★ 6.6k754 structured cybersecurity skills for AI agents · Mapped to 5 frameworks: MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, D3FEND & NIST AI RMF · agentskills.io standard · Works with Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI & 20+ platform
Hooker
★ 5.1k🔥🔥 hooker is a Frida-based reverse engineering toolkit for Android. It offers a user-friendly CLI, universal scripts, auto hook generation, memory roaming to detect activities/services, one-click SOCKS5 proxy setup, Frida JustTrustMe, and BoringSSL u
Browse More Security MCP Servers
Explore all security servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Medusa in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Medusa?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.