MCP servers · by job

Security MCP servers

119 MCP servers that scan, test and investigate security issues. 31 are SAFE, 21 CAUTION and 67 BLOCK. Ranked by trust score; each card opens the full audit.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “Security MCP servers”. Retrieved 28 Oct 2026. https://modelsatlas.com/mcp/topic/security

Verdicts · 119 servers

✓SAFE31!CAUTION21✕BLOCK67

Transports

A server can list more than one transport.

Tool risk

26

servers expose at least one destructive tool: delete, overwrite or run. The median server with a readable tool list exposes 14 tools.

Servers with no destructive tools →

119 security servers

1–24 of 119 · ranked by trust score, then stars

Agentic Radar

splx-ai/agentic-radar

✓SAFE B

A security scanner for your LLM agentic workflows

★ 1,0531 writesse

Kali Connector

wh0am123/mcp-kali-server

✓SAFE B

MCP configuration to connect AI agent to a Linux machine.

★ 8302 write

Security Hub

fuzzinglabs/mcp-security-hub

✓SAFE B

A growing collection of MCP servers bringing offensive security tools to AI assistants. Nmap, Ghidra, Nuclei, SQLMap, Hashcat and more.

★ 7948 writestdio

Apktool

zinja-coder/apktool-mcp-server

!CAUTION B

A MCP Server for APK Tool (Part of Android Reverse Engineering MCP Suites)

★ 6592 writestreamable-http

H1 Brain

patrikfehrenbach/h1-brain

✓SAFE B

MCP server that connects AI assistants to HackerOne for bug bounty hunting

★ 3553 write

OnionClaw

christinminor459/onionclaw

!CAUTION B

Provide AI agents with full Tor network access and dark web data through a zero-config OpenClaw skill or standalone tool.

★ 3411 write

Maigret

burtthecoder/mcp-maigret

✓SAFE B

MCP server for maigret, a powerful OSINT tool that collects user account information from various public sources.

★ 2672 readstdio

OpenClaw

freema/openclaw-mcp

!CAUTION B

🦞 MCP server for OpenClaw - secure bridge between Claude.ai and your self-hosted OpenClaw assistant with OAuth2 authentication

★ 1852 writesse

Shodan

burtthecoder/mcp-shodan

✓SAFE B

MCP server for Shodan — search internet-connected devices, IP reconnaissance, DNS lookups, and CVE/CPE vulnerability intelligence. Works with Claude Code, Codex, Gemini CLI, and Claude Desktop.

★ 1747 readstdio

VirusTotal

burtthecoder/mcp-virustotal

✓SAFE B

MCP server for VirusTotal API — analyze URLs, files, IPs, and domains with comprehensive security reports, relationship analysis, and pagination support.

★ 1505 readstdio

x64dbg

bromoket/x64dbg_mcp

✓SAFE B

Drive x64dbg with your AI. MCP server: 23 mega-tools / 153 endpoints for breakpoints, memory, disasm, tracing, anti-debug & PE dumping. Claude/Cursor/Windsurf/Cline. All local.

★ 1318 writestdio

Cve Search

roadwy/cve-search_mcp

✓SAFE B

A Model Context Protocol (MCP) server for querying the CVE-Search API

★ 1071 writesse

Awsome Kali

ccq1/awsome_kali_mcpservers

✓SAFE B

awsome kali MCPServers is a set of MCP servers tailored for Kali Linux

★ 10627 readstdio

ExternalAttacker

mordavid/externalattacker-mcp

✓SAFE B

A modular external attack surface mapping tool integrating tools for automated reconnaissance and bug bounty workflows.

★ 778 readstdio

Cheat Engine

bethington/cheat-engine-server-python

✓SAFE B

MCP Cheat Engine Server — provides safe, structured read-only access to memory analysis and debugging functionality through the Model Context Protocol (MCP). For developers, security researchers, and game modders.

★ 664 destructive

OSINT

badchars/osint-mcp-server

✓SAFE B

OSINT intelligence MCP server for AI agents — 37 tools, 12 sources. Shodan, VirusTotal, Censys, SecurityTrails, DNS reconnaissance, WHOIS, certificate transparency, BGP routing, Wayback Machine, GeoIP. Automated open source intelligence and attack surface mapping via Model Context Protocol.

★ 601 writestdio

Security Audit

qianniuspace/mcp-security-audit

✓SAFE B

A powerful MCP (Model Context Protocol) Server that audits npm package dependencies for security vulnerabilities. Built with remote npm registry integration for real-time security checks.

★ 571 readstdio

DNStwist

burtthecoder/mcp-dnstwist

✓SAFE B

MCP server for dnstwist, a powerful DNS fuzzing tool that helps detect typosquatting, phishing, and corporate espionage.

★ 511 readstdio

Shodan

cyreslab-ai/shodan-mcp-server

✓SAFE B

MCP server for Shodan — host/network recon, CVE and exploit lookup, InternetDB, and persistent alerts (34 tools).

★ 502 destructivestdio

Copilot Security

robotti-io/copilot-security-instructions

✓SAFE B

✨ A customizable copilot-instructions.md ruleset & prompts to guide GitHub Copilot toward secure coding defaults in Java, Node.js, C# and Python. Blocks risky patterns, teaches safe habits.

★ 423 readstreamable-http

ExploitDB

cyreslab-ai/exploitdb-mcp-server

✓SAFE B

MCP server for Exploit-DB — search public exploits and shellcode by CVE, platform, type, and tags.

★ 3118 readstdio

MalwareBazaar

mytechnotalent/malwarebazaar_mcp

✓SAFE B

An AI-driven MCP server that autonomously interfaces with Malware Bazaar, delivering real-time threat intel and sample metadata for authorized cybersecurity research workflows.

★ 304 readstdio

Binary Analysis

ap3x/binaryanalysis-mcp

✓SAFE B

MCP server for analyzing PE, ELF, and Mach-O binaries using LIEF

★ 259 readstdio

Kali Go

found-cake/kali-mcp-go

✓SAFE B

MCP runtime for AI-assisted security testing with Kali tools, explicit target controls, bounded scanning, structured evidence, and persistent Docker deployment.

★ 253 read

Other jobs

By risk

Questions

How do you decide which servers are security servers?
By the categories each server's own repository declares, mapped onto this job by a hand-kept table. A server can belong to more than one job, and one whose repository declares no matching category is not listed here.
Which security MCP servers rank highest on trust?
Agentic Radar (splx-ai/agentic-radar), Kali Connector (wh0am123/mcp-kali-server) and Security Hub (fuzzinglabs/mcp-security-hub), by trust score and then stars. Ranking never uses popularity to change a verdict.
How many security servers only read?
35 of the 119 declare tools that all only read; 26 expose at least one destructive tool.

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.