Security MCP servers for scanning, threat intelligence and audits: let an AI assistant check code, dependencies and infrastructure for risks. 3,538 servers in this category.
Judges AI agent proposals against constitutional policy floors before execution and records a sealed verdict receipt.
$claude mcp add arifos -- uvx arifos
Performs black-box security assessments on live MCP servers by sending realistic and malformed inputs to identify vulnerabilities.
Gives AI coding agents a local SQLite memory plus an Ed25519 and post-quantum signed, offline-verifiable audit trail.
$claude mcp add world-model-mcp -- uvx world-model-mcp
Exposes the reconFTW reconnaissance framework through MCP so AI assistants can perform security scans and inspect results.
Enables AI assistants to perform automated binary diffing by comparing IDA Pro databases and inspecting function changes and call graphs.
An OSINT investigator's toolkit that lets an AI client perform IP reputation lookups, domain intelligence, email header analysis, and phishing hunting.
Looks up IP addresses to return geolocation, ASN, ISP, and proxy details for AI clients.
$claude mcp add mcp-ip2location-io -- uvx mcp-ip2location-io
Acts as a security gateway for MCP agents to inspect tool calls, enforce policies, and block risky filesystem, command, or network operations in real time.
Turns declarative YAML configurations into secure, policy-enforced MCP tool servers with built-in trust enforcement and audit logging.
Provides a secure authentication proxy for MCP servers with centralized OAuth 2.0, JWT authorization, and real-time monitoring.
Scans codebases to detect AI frameworks and generate compliance reports for the EU AI Act and GDPR.
$claude mcp add --transport http mcp-eu-ai-act https://mcp.arkforge.tech/mcp
Enables autonomous AI agents to perform workflows within isolated Qubes OS sandboxes using dom0-mediated security boundaries.
Performs static security scanning for MCP servers, AI agent skills, and plugins using 17 attack patterns and 65 regex signatures.
$claude mcp add skill-audit-mcp -- docker run -i --rm ghcr.io/eltociear/skill-audit-mcp:mcp-1.2.0
Acts as an independent admission controller for AI systems by aggregating code risk, evaluations, and traces into auditable release decisions.
$claude mcp add release-gate -- uvx release-gate
Provides a security layer for AI agents featuring scoped permissions, encrypted secret storage, and tamper-evident audit logging.
$claude mcp add haldir -- uvx haldir
Provides a security-hardened MCP interface for Kali Linux using OAuth 2.0, TLS, and strict tool allowlisting to prevent arbitrary command execution.
An MCP control plane that manages security scanning by deduplicating results, enforcing rate limits, and running non-destructive proofs.
$claude mcp add keel -- uvx keel-pentest
Performs AI-powered code reviews with local security-pattern checks, N+1 detection, and custom team rule enforcement.
$claude mcp add aicraft -- uvx aicraft-code-review
Provides LLMs with software composition analysis capabilities including license compliance, binary analysis, and vulnerability assessment.
Translates between all trust credential formats used by AI agents via a canonical Universal Trust Schema.
$claude mcp add universal-trust-adapter -- npx -y marketnow-mcp