Skip to content

MCP servers for security

89 MCP servers that scan for vulnerabilities, manage secrets and investigate threats. Each one shows how to install it in Claude, Cursor and VS Code, whether it runs locally or remotely, and the tools it provides.
  1. Score AI agent runs for quality, PII leaks, prompt injection and cost, on your own machine

    Runs locally
    • 3 tools
    • Open source
    • AI models and LLM tooling
  2. Install and run open-source OSINT tools in a local sandbox, asking before target runs

    Runs locally
    • 8 tools
    • Open source
    • Security
  3. Write, test, lint and explain OPA Rego policies from any MCP client

    Runs locally
    • 10 tools
    • Open source
    • Developer tools
  4. Inspect and manage a UniFi network: devices, clients, WiFi, firewall and VPN

    Runs locally
    • Open source
    • Cloud and infrastructure
  5. Read-only search of an OSINT tool and guide registry, with consent classes for each tool

    Not checked yet
    • 6 tools
    • Remote
    • Free
    • Security
  6. Secret broker that injects credentials at the point of use, never into the model

    Runs locally
    • Open source
    • Security
  7. Let a coding agent use temporary credentials without the values entering the chat

    Runs locally
    • Free
    • Security
  8. Journal, approve or block, and sandbox every MCP tool call your AI agent makes

    Runs locally
    • Open source
    • Security
  9. Local checks for sudo, systemd, sshd, ufw and secret leaks before an agent runs commands

    Not checked yet
    • 3 tools
    • Remote
    • Open source
    • Security
  10. Let two companies' AI agents work together on a signed record, with people approving

    Not checked yet
    • Remote
    • Free
    • Communication
  11. Self-hosted customer feedback search, stats and themes, with PII redacted at ingest

    Runs locally
    • Open source
    • Analytics and data
  12. Read your Golem agent fleet sessions, live activity and signed record from any client

    Not checked yet
    • Remote
    • Free
    • Security
  13. Offline security inspection for MCP configs, metadata and dependencies, run from the CLI

    Runs locally
    • Open source
    • Security
  14. Pins MCP tool definitions in a lockfile and flags rug-pulls, poisoning and leaked secrets

    Runs locally
    • 2 tools
    • Open source
    • Security
  15. Read-only MCP server that lints and budgets CLAUDE.md, AGENTS.md and other steering files

    Runs locally
    • Open source
    • Files and documents
  16. Check whether your exit IP works for ChatGPT, Claude and Gemini from your AI client

    Runs locally
    • 2 tools
    • Open source
    • AI models and LLM tooling
  17. Consent-based, end-to-end encrypted messaging between AI agents with a public ledger

    Runs locally
    • 9 tools
    • Open source
    • Communication
  18. Offline secret scanning MCP server that returns redacted findings to coding agents

    Runs locally
    • 2 tools
    • Open source
    • Security
  19. macOS Keychain secret broker where an agent can use your API keys but never read them

    Runs locally
    • 6 tools
    • Open source
    • Security
  20. Read-only security scanner for code, infrastructure files and AI agent config

    Runs locally
    • Free
    • Security
  21. Encrypted local vault where your AI fetches API keys on demand and scans code for leaks

    Runs locally
    • 17 tools
    • Free
    • Security
  22. Local approvals and audit so coding agents use credentials without ever seeing them

    Runs locally
    • Free
    • Security
  23. Network scanner for AI agents: host discovery, port scans, DNS, ARP and mDNS lookups

    Runs locally
    • 13 tools
    • Open source
    • Security
  24. Let an agent clear node_modules and other junk folders, only inside roots you allow

    Runs locally
    • 1 tool
    • Open source
    • Developer tools
  25. Read-only npm preflight: reviews what install scripts do before packages run

    Runs locally
    • Open source
    • Security
  26. Tamper-evident local record of what your AI agent did, plus verifiable markdown memory

    Runs locally
    • 9 tools
    • Open source
    • Knowledge and memory
  27. Find and strip zero-width, bidi and tag characters hidden in text, offline

    Runs locally
    • 5 tools
    • Open source
    • Security
  28. Access control for AI agents: allow, hold for approval or block each tool call

    Runs locally
    • Open source
    • Security
  29. MCP server and open protocol for signed messages between agents with human-gated approvals

    Runs locally
    • Open source
    • Developer tools

See every kind of AI for security: tools, agents, MCP servers and skills.Ranked: the best AI for security.