Security MCP servers for scanning, threat intelligence and audits: let an AI assistant check code, dependencies and infrastructure for risks. 3,537 servers in this category.
Transforms Claude into a cyber threat intelligence and OSINT analyst using structured collection, automated pivots, and intelligence reporting.
Verifies security for AI model artifacts, live endpoints, MCP servers, and recorded agent traces using built-in security rules.
Screens an AI agent's memory writes and reads through detectors and a declarative policy to block poisoning attacks.
Lets an AI client manage a pfSense firewall in natural language: rules, NAT, VPN, DNS, DHCP, certificates and diagnostics, with safety guardrails.
Lets an AI client verify intent, provenance, and signed evidence before executing high-impact tool calls like payments or data deletion.
Lets an AI client store, search, and verify red-team credentials such as API keys, cloud AK/SK, and SSH keys through a local encrypted vault.
Lets AI agents look up CVEs, KEV entries, CWE weaknesses, MITRE techniques, Sigma rules, and domain or IP security posture.
$claude mcp add --transport http contrastapi https://api.contrastcyber.com/mcp/
Provides safety and decision-making tools for coding agents, including command validation, pull request routing, and context management.
Lets AI assistants fetch and read Gopher and Gemini protocol resources, including menus, gemtext pages, and index searches.
$claude mcp add gopher-mcp -- uvx gopher-mcp
Capture and inspect HTTP requests, emails, and DNS lookups to test webhooks, debug API callbacks, and perform security testing.
$claude mcp add webhook-mcp-server -- uvx webhook-mcp-server
Enables AI agents to obfuscate Python projects, scan for obfuscation risks, generate configuration files, and reverse obfuscated stack traces.
$claude mcp add pyobfus-mcp -- uvx pyobfus-mcp
Scans text, files, and tool calls for security threats and credential leaks to act as an input firewall for AI agents.
$claude mcp add sunglasses -- uvx sunglasses
Enables an AI client to perform read-only verification of autonomous agent evidence packages using three-valued verdicts and Ed25519 seals.
$claude mcp add uibc-core -- uvx uibc-core
Acts as a stdio proxy between MCP clients and tool servers to enforce security contracts and prevent unauthorized tool calls based on untrusted output.
$claude mcp add mcp-bouncer -- uvx bouncer-mcp
Enables AI clients to query threat intelligence, reason over adversary TTPs, and generate ATT&CK Navigator layers using official STIX data.
$claude mcp add complete-mitre-attack-mcp-server -- npx -y @imouiche/mitre-attack-mcp-server
Connect AI clients to MCP servers through a secure, centralized management layer that handles credentials and profiles locally.
$claude mcp add lightnow-proxy -- uvx lightnow-proxy
Provides self-hosted code intelligence for cross-repository Q&A, dependency audits, and automated pull request reviews via MCP.
$claude mcp add celmis -- docker run -i --rm ghcr.io/celmis-labs/celmis-api:v0.1.30
Enables AI agents to research threat intelligence from AlienVault OTX, including pulses, indicators, reputation, and malware analysis.
$claude mcp add mcp-otx -- uvx otx-mcp
Audits AI agent skills using static, semantic, adversarial, and supply-chain scanning to identify security risks and malicious intent.
$claude mcp add tar-engine -- uvx tar-engine
Provides a pre-payment safety gate for agents to screen PII, enforce spending policies, and detect duplicate payments before execution.
$claude mcp add presidio-hardened-x402-mcp -- uvx presidio-hardened-x402-mcp