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.
Provides a secure SQL proxy and credential vault that allows AI agents to query databases using natural language with AST-level safety and row-level security.
$claude mcp add queryshield -- uvx queryshield-mcp
Prevents AI agents from leaking private repository data to public repositories by enforcing flow guards, human approvals, and audit logs.
$claude mcp add repoward -- uvx repoward
Provides a production-oriented OAuth 2.1 and OIDC resource-server reference for securing MCP tools with Microsoft Entra ID or generic OIDC.
$claude mcp add mcp-server-auth-template -- docker run -i --rm ghcr.io/brunovicco/mcp-server-auth-template:v0.7.0
Signs, hash-chains, and verifies cryptographically signed compute usage receipts across any provider.
$claude mcp add computeledger -- uvx computeledger-cli
Lets an AI client capture, query, group, and inspect Charles Proxy live and recorded HTTP traffic for analysis.
Lets an AI client search .onion sites, fetch hidden services through Tor, rotate its identity, and run OSINT analysis pipelines.
Runs 135 Kali Linux security tools in a container so an AI client can scan, fuzz, enumerate and exploit targets.
Reduces AI coding agent token usage and prevents security leaks by providing focused context through AST slicing and local guardrails.
$claude mcp add tokensafe -- npx -y tokensafe
Scans Model Context Protocol servers for RCE, path traversal, prompt injection, and protocol vulnerabilities, producing JSON, HTML, or SARIF reports.
Lets AI agents control side-channel and fault-injection hardware to capture power traces, perform CPA, and sweep glitch parameters.
Connects AI agents to AuditHub so they can browse organizations and projects, launch and monitor security audit tasks, and read issues and findings.
Enables AI agents to perform penetration testing, vulnerability assessment, reconnaissance, and security reporting through natural language.
Lets AI agents search and fetch Tor/.onion sites through 17 dark web engines with LLM-powered OSINT analysis.
Provides a secure execution infrastructure for AI agents that decouples agent logic from sensitive credentials and LLM providers.
$claude mcp add --transport http identark https://api.identark.io/v1/mcp/rpc
Connects AI coding assistants to Snyk API & Web to onboard scan targets, run DAST scans, and triage security findings using natural language.
$claude mcp add saw-mcp -- uvx snyk-apiweb-mcp
Provides a multi-tenant platform to connect AI assistants to databases and filesystems through a single authenticated endpoint with RBAC and audit logging.
$claude mcp add --transport http mcp-gateway https://your-gateway.example.com/t/your-org/mcp/sse
Equip AI agents with a secure toolkit featuring a prompt firewall, 74 specialized tools, persistent memory, and web search capabilities.
Allows an AI client to investigate and manage NSX/vDefend security policies, firewall rules, groups, and troubleshooting data.
$claude mcp add vmware-nsx-security -- uvx vmware-nsx-securityPerforms VMware compliance scanning against CIS, vSphere SCG, and PCI-DSS baselines while tracking configuration drift.
$claude mcp add vmware-harden -- uvx vmware-harden
Enforces strict runtime budgets and capability-based security for code generated by AI models.
$claude mcp add velaris-lang -- uvx velaris-lang