MCP Scan
Offensive MCP server auditor — detects tool poisoning, credential leaks, RCE vectors, SSRF, session hijacking, and supply chain vulnerabilities across stdio, HTTP, and SSE transports.
What is MCP Scan?
MCP Scan is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to offensive mcp server auditor — detects tool poisoning, credential leaks, rce vectors, ssrf, session hijacking, and supply chain vulnerabilities across stdio, http, and sse transports.
Offensive MCP server auditor — detects tool poisoning, credential leaks, RCE vectors, SSRF, session hijacking, and supply chain vulnerabilities across stdio, HTTP, and SSE transports.
This server falls under the Security category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Offensive MCP server auditor — detects tool poisoning, crede
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx mcpscanConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use MCP Scan
MCPScan is a TypeScript-based offensive security auditing tool for MCP servers that detects tool poisoning, credential leaks, RCE vectors, SSRF, session hijacking, overprivileged tools, missing authentication, and supply chain vulnerabilities across stdio, HTTP, and SSE transports. It auto-discovers MCP configurations from Claude Desktop, Cursor, and other common config locations, and can also scan specific server commands or live HTTP endpoints. MCPScan checks against 8 vulnerability categories and references known CVEs including high-severity issues in mcp-remote, MCP Inspector, and Claude Code itself.
Prerequisites
- Node.js 18 or later
- npm for building the project
- Git to clone the repository
- MCP server configurations already set up (claude_desktop_config.json, .cursor/mcp.json, or similar) for the auto-discovery scan mode
Clone the repository
Clone the MCPScan repository to your local machine.
git clone https://github.com/sahiloj/MCPScan.git
cd MCPScanInstall dependencies and build
Install Node.js dependencies and compile the TypeScript source to create the mcpscan binary.
npm install
npm run buildLink the CLI globally
Create a global symlink so mcpscan is available as a CLI command from any directory.
npm linkScan all auto-discovered MCP configs
Run mcpscan with --all-configs to automatically find and audit MCP server configurations from Claude Desktop, Cursor, and other common locations on your system.
mcpscan scan --all-configsScan a specific server command or HTTP endpoint
Audit a single MCP server by specifying its command and arguments, or point at a live HTTP endpoint. Optionally limit which vulnerability checks to run.
# Scan a specific stdio server
mcpscan scan --command "npx" --args "-y @modelcontextprotocol/server-filesystem /home/user"
# Scan an HTTP/SSE endpoint
mcpscan scan --target http://localhost:3000/mcp
# Run only specific checks
mcpscan scan --all-configs --checks tool-poisoning,credential-leak,rce-vectors
# Output SARIF format for CI/CD integration
mcpscan scan --all-configs --output sarif > findings.sarifMCP Scan Examples
Client configuration
MCPScan is a CLI security tool, not a server you connect a client to. Configure it as an MCP server if you want Claude to be able to invoke scans on demand.
{
"mcpServers": {
"mcpscan": {
"command": "mcpscan",
"args": ["serve"]
}
}
}Common scan commands
Practical MCPScan CLI invocations for auditing MCP server configurations and endpoints.
# Full audit of all detected MCP configs
mcpscan scan --all-configs
# Audit a specific server package
mcpscan scan --command "npx" --args "-y @modelcontextprotocol/server-filesystem /home/user"
# Scan live HTTP MCP endpoint
mcpscan scan --target http://localhost:3000/mcp
# Focused checks only
mcpscan scan --all-configs --checks tool-poisoning,credential-leak,rce-vectors
# SARIF output for GitHub Advanced Security or CI
mcpscan scan --all-configs --output sarif > findings.sarifTroubleshooting MCP Scan
mcpscan command not found after npm link
Run 'npm run build' first to generate the dist/ output before linking. Verify with 'which mcpscan'. If still not found, check that your npm global bin directory is on your PATH with 'npm bin -g'.
Auto-discovery finds no configs with --all-configs
MCPScan checks platform-specific paths: ~/Library/Application Support/Claude/claude_desktop_config.json on macOS, ~/.config/claude/claude_desktop_config.json on Linux, %APPDATA%\Claude\claude_desktop_config.json on Windows, and .cursor/mcp.json. Ensure at least one of these files exists with mcpServers configured.
Scan reports false positives for credential-leak checks
MCPScan performs static analysis of tool descriptions and parameter names for patterns matching AWS keys, JWTs, API tokens, and database connection strings. Review flagged items manually — pattern-matched credential detection may flag legitimate parameter names. Use --checks to exclude categories not relevant to your threat model.
Frequently Asked Questions about MCP Scan
What is MCP Scan?
MCP Scan is a Model Context Protocol (MCP) server that offensive mcp server auditor — detects tool poisoning, credential leaks, rce vectors, ssrf, session hijacking, and supply chain vulnerabilities across stdio, http, and sse transports. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install MCP Scan?
Follow the installation instructions on the MCP Scan GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with MCP Scan?
MCP Scan works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is MCP Scan free to use?
Yes, MCP Scan is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
MCP Scan Alternatives — Similar Security Servers
Looking for alternatives to MCP Scan? 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 MCP Scan 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 MCP Scan?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.