MCP Armor

v1.0.0Securitystable

MCP Armor continuously secures and monitors Model Context Protocol operations through static and dynamic scans, revealing hidden risks in agent-to-tool communications.

agentic-aiaiai-agentsairamcp
Share:
115
Stars
0
Downloads
0
Weekly
0/5

What is MCP Armor?

MCP Armor is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to mcp armor continuously secures and monitors model context protocol operations through static and dynamic scans, revealing hidden risks in agent-to-tool communications.

MCP Armor continuously secures and monitors Model Context Protocol operations through static and dynamic scans, revealing hidden risks in agent-to-tool communications.

This server falls under the Security category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • MCP Armor continuously secures and monitors Model Context Pr

Use Cases

Agent-to-tool communication security
Static and dynamic vulnerability scanning
MCP operation monitoring
aira-security

Maintainer

LicenseApache-2.0
Languagepython
Versionv1.0.0
UpdatedMay 14, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx mcp-armor

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use MCP Armor

MCP Armor is a local security scanner for Model Context Protocol deployments that performs both static and dynamic analysis of MCP server configurations to detect vulnerabilities in agent-to-tool communications. It can identify prompt injection risks, rug pull attacks, cross-server tool shadowing, tool poisoning, command injection vulnerabilities, hardcoded secrets, and baseline drift from unauthorized changes. Security engineers and teams running multi-server MCP deployments can use MCP Armor to audit their configurations before deploying AI agents or to continuously monitor for unexpected changes.

Prerequisites

  • Python 3.9+ and pip installed
  • One or more MCP server configuration files (e.g., claude_desktop_config.json or similar) to scan
  • MCP Armor runs entirely locally — no cloud account or API key is required
  • An MCP-compatible client if you want to expose MCP Armor as an MCP tool itself (optional)
1

Install MCP Armor via pip

Install the mcp-armor package from PyPI. A virtual environment is recommended to isolate dependencies.

python3 -m venv .venv
source .venv/bin/activate
pip install mcp-armor
2

Create a baseline of your current MCP configuration

Run the inspect command to catalog all tools, resources, and prompts currently exposed by your MCP servers. This creates a baseline file used for drift detection in subsequent scans.

mcp-armor inspect --config ~/.config/Claude/claude_desktop_config.json --baseline ./mcp-baseline.json
3

Run a security scan against your configuration

Run the scan command to perform static security analysis. MCP Armor checks for prompt injection, tool poisoning, cross-server shadowing, hardcoded secrets, and command injection vulnerabilities.

mcp-armor scan --config ~/.config/Claude/claude_desktop_config.json --baseline ./mcp-baseline.json --report-type md --output ./mcp-security-report.md
4

Review the security report

Open the generated report file (Markdown or JSON format). Each finding includes the vulnerability type, affected tool or resource, severity, and recommended remediation steps.

# View the markdown report
cat ./mcp-security-report.md
5

Integrate into CI or run periodically

Add MCP Armor to your CI pipeline or run it on a schedule to detect drift from the baseline. Use --verbose for detailed terminal output during automated runs.

mcp-armor scan --config ./claude_desktop_config.json --baseline ./mcp-baseline.json --report-type json --output ./report.json --verbose

MCP Armor Examples

Client configuration

Example MCP Armor scan command covering multiple MCP config files simultaneously, outputting a JSON report for automated processing.

{
  "mcpServers": {
    "mcp-armor": {
      "command": "mcp-armor",
      "args": ["scan",
        "--config", "/path/to/claude_desktop_config.json",
        "--baseline", "/path/to/mcp-baseline.json",
        "--report-type", "json",
        "--output", "/path/to/report.json"
      ]
    }
  }
}

Prompts to try

CLI commands and use cases to explore MCP Armor's security scanning capabilities.

- mcp-armor inspect --config claude_desktop_config.json --baseline baseline.json
- mcp-armor scan --config claude_desktop_config.json --baseline baseline.json --report-type md
- mcp-armor scan --config config1.json --config config2.json --verbose
- mcp-armor scan --config claude_desktop_config.json --show-logs
- Review the report for 'Prompt Injection' and 'Tool Poisoning' findings
- Compare two baseline files to identify configuration drift over time

Troubleshooting MCP Armor

mcp-armor command not found after pip install

Ensure the virtual environment is activated with 'source .venv/bin/activate' before running mcp-armor. Alternatively, install with 'pip install --user mcp-armor' and ensure ~/.local/bin is in your PATH.

Inspect or scan fails with 'config file not found' error

Provide the absolute path to your MCP configuration file with the --config flag. For Claude Desktop on macOS, the config is at '~/Library/Application Support/Claude/claude_desktop_config.json'. Use 'ls' to verify the file exists before running the scan.

Scan reports false positives for known-safe tools

Update your baseline file by re-running 'mcp-armor inspect' after reviewing and confirming the tools are safe. The drift detection compares against the baseline, so an up-to-date baseline reduces false drift alerts in subsequent scans.

Frequently Asked Questions about MCP Armor

What is MCP Armor?

MCP Armor is a Model Context Protocol (MCP) server that mcp armor continuously secures and monitors model context protocol operations through static and dynamic scans, revealing hidden risks in agent-to-tool communications. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install MCP Armor?

Follow the installation instructions on the MCP Armor GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with MCP Armor?

MCP Armor works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is MCP Armor free to use?

Yes, MCP Armor is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.

MCP Armor Alternatives — Similar Security Servers

Looking for alternatives to MCP Armor? Here are other popular security servers you can use with Claude, Cursor, and VS Code.

Casdoor

13.6k

An 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

9.0k

An 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

8.9k

HexStrike 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

8.7k

Enables 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

6.6k

754 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.

Quick Config Preview

{ "mcpServers": { "mcp-armor": { "command": "npx", "args": ["-y", "mcp-armor"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use MCP Armor?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides