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.
Enforces deterministic data governance at the AI agent boundary through runtime validation, maturity scoring, and cryptographically signed S-Certs.
$claude mcp add odgs-mcp-server -- uvx odgs-mcp-server
Decode, inspect, modify, and rebuild Android APK projects using APKTool directly from an AI client.
Lets an AI client import vulnerability scanner results and run nmap, NetExec, nuclei, or Metasploit against affected hosts to verify findings.
Provides an agentic memory system for cyber threat intelligence that extracts entities, builds STIX 2.1 knowledge graphs, and resolves threat actor aliases.
Adds PHI redaction, immutable audit, step-up auth, and tenant isolation between AI agents and FHIR clinical data.
$claude mcp add --transport http healthclawguardrails https://mcp-server-production-5112.up.railway.app/mcp
An MCP gateway that provides governed Postgres access for AI assistants using OIDC authentication, row-level security, and human-approved writes.
Lets an AI client plan, inspect, and approve shell commands in a scoped workspace, then run them under limits and store signed receipts.
Acts as a provenance gateway for MCP servers to block tool calls triggered by untrusted or malicious content.
$claude mcp add reasongate -- uvx reasongate
Provides AI agents with WHOIS/RDAP, DNS, SSL, subdomain, and email security lookups for any domain.
$claude mcp add --transport http domain-intelligence https://oti-labs.com/mcp
Allows AI assistants to query and analyze HPE GreenLake audit logs by filtering by category, user, and timestamp to track system events.
$claude mcp add gl-mcp -- uvx greenlake-audit-logs-mcp
Performs repository-wide static analysis to identify security vulnerabilities, defects, and architectural dependencies with cited evidence.
Enriches LLM context with real-time threat intelligence by checking domains, URLs, and IP addresses against ten security blacklists.
A cryptographic gateway that allows LLMs to perform aggregate queries over encrypted data without accessing individual rows.
Statically scans MCP tool manifests and runs prompt-injection attack suites to detect security vulnerabilities like tool poisoning and jailbreaks.
$claude mcp add sentinel-scan-cli -- npx -y sentinel-scan-cli
Provides AI assistants with PyPI package metadata and vulnerability auditing against the OSV database.
Allows an AI client to perform dry-run policy evaluations for agent security without live enforcement.
$claude mcp add vantio-open-core -- npx -y @vantio/gate-mcp
Provides AI agents with read-only CVE intelligence, remediation playbooks, and evidence-gated vulnerability remediation workflows.
$claude mcp add --transport http security-recipes-ai https://security-recipes.ai/mcp
Acts as a privacy and approval gateway that enforces human oversight and PII detection for AI assistants accessing business systems.
$https://github.com/privacyfence/privacyfence/releases/download/v7.0.2/PrivacyFence.mcpb
Enforces permission tiers, approval gates, and audit logging directly within MCP servers to control AI tool access.
$claude mcp add mcp-gatehouse -- uvx mcp-gatehouse
Enables AI agents to perform threat hunting and traffic analysis by querying Malcolm instances including Zeek, Suricata, Arkime, and OpenSearch.
$claude mcp add mcp-server-malcolm -- uvx mcp-server-malcolm