190 MCP servers and agent skills in the Security category, ranked by quality score.
Human sign-off + trust receipts for AI agents: requires a named human's approval before an irreversible action (payment release, record change, deploy), then mints an offline-verifiable Ed25519 Trust Receipt. Also exposes trust profiles, receipt verification, disputes, and delegation. Apache-2.0; policy engine formally verified. Install: `npx -y @emilia-protocol/mcp-server`.
[Mobb Vibe Shield](https://vibe.mobb.ai/) identifies and remediates vulnerabilities in both human and AI-written code, ensuring your applications remain secure without slowing development.
Local-first, zero-egress security scanner for AI-generated / "vibe-coded" JS/TS. Bundles Opengrep, Gitleaks & Trivy behind one CWE-keyed schema and adds AI-code-specific checks (client-side secret exposure, Supabase RLS, prompt-injection & LLM-output XSS sinks). No account, no telemetry.
Proof-of-behavior enforcement for AI agents. Define behavioral covenant rules (permit/forbid/require), enforce at runtime before execution, get SHA-256 hash-chained tamper-evident audit logs, and verify compliance independently. Cross-agent verification handshake — no proof, no transaction. MIT licensed, 4,244 tests.
AI agent identity, trust, and audit trail infrastructure. 17 MCP tools: register agents with Ed25519 keypairs, check permissions (sub-5ms), emit tamper-evident audit events, verify trust scores (0-100), delegate credentials, and manage ephemeral agents. IETF Internet-Draft filed (draft-vorim-vaip-00). Works with LangChain, OpenAI, CrewAI, Stripe ACP, and 4 more frameworks. `npx @vorim/mcp-server`.
MCP server for querying the VirusTotal API. This server provides tools for scanning URLs, analyzing file hashes, and retrieving IP address reports.
Human-in-the-loop approval inbox for AI agents. An agent submits a proposed action (send email, post comment, run a command) via `impri_push_action`, a human approves, rejects, or edits it from a web, mobile, or Slack/Discord/Telegram inbox, and the agent only proceeds on an approved decision. The gate is a data dependency, not a prompt. Full audit trail, self-hostable (MIT, Docker Compose). `npx
AI Agent Security Middleware & MCP Server with 8-layer defense including prompt injection detection, DLP data flow tracking, command blocking, and PII detection. 7 MCP tools, zero dependencies.
MCP Server for managing & interacting with Open Source NGFW OPNSense via Natural Language
Compliance and identity for AI agents — verify an agent's certificate scope, read its behavioral risk score, and append to an immutable audit trail. X.509 identity issued via AWS KMS; MiCA / EU AI Act aligned. `npx -y @kakunin/mcp`
Agent identity, scoped delegation, and governance: issue passports, build and verify narrowing-only delegation chains, enforce policy at a gateway, and emit signed admission and outcome records. 150 tools. `npx -y agent-passport-system-mcp`
Local credential approval and execution gateway for AI coding agents. Agents receive typed handles instead of raw API tokens, SSH keys, and cloud credentials; one-time approvals bind the credential, command, arguments, working directory, environment bindings, and target. Injects credentials only into the approved child process, sanitizes output, and records a local audit trail. Install with `npm i
Governance proxy for MCP servers. Wraps any upstream server with policy evaluation, human approval workflows, and hash-chain audit trails. 18+ framework integrations. Apache 2.0 SDK.
Credential isolation proxy for AI agents. Injects secrets at the network boundary with domain restrictions, agent authentication, and audit logging. No SDK required — works as a transparent HTTP proxy or MCP server.
MCP server for Bitwarden and Vaultwarden vault management. Search, create, edit, and organize logins, notes, cards, identities, SSH keys, folders, collections, attachments, and Sends via the official `bw` CLI.
Security-first MCP governance proxy (`sint-mcp`) with capability tokens, T0-T3 approval tiers, fail-closed execution, and tamper-evident audit receipts. Includes a separate `sint-scan` CLI for preflight MCP tool-risk audits.
Post-quantum readiness for AI coding agents: scan code for quantum-vulnerable cryptography (RSA/ECDH/ECDSA/DH), explain the harvest-now-decrypt-later exposure, get NIST ML-KEM/ML-DSA/SLH-DSA (and hybrid) migration guidance, verify fixes, and check dependencies. Content-based/advisory tools only. Run local (`npx @quantakrypto/mcp`) or the hosted OAuth endpoint at [mcp.quantakrypto.com](https://mcp.
ARGUS-3 as a stdio MCP server** (`argus mcp` → `argus_ask`, `argus_status`). **WARDEN** vets third-party MCP servers before any tool runs (LUMEN-scored firewall, tool-def pinning, drift sentinel). Distinct from `aimarket-oracle-gateway` (oracle tools) and `aimarket-plugins` (hub packager). npm `@alexar76/argus3` · [live](https://magic-ai-factory.com/argus/).
Open Policy Agent (OPA) and Rego policy toolkit. 32 tools spanning authoring (format, lint, check, deps), evaluation (eval, test, bench, coverage), and OPA REST control (policies, data, decisions, compile). Wraps the OPA CLI and the [Regal](https://github.com/StyraInc/regal) linter, with AI-assisted helpers for explaining decisions, generating test skeletons, and suggesting fixes.
Injects a live, reversible MCP server into a running Express / FastAPI / Next.js app — reads safe by default, writes gated behind human confirmation. The same engine also proves deploys and PRs (`apocalypse` / `review`).
MCP server for maigret, a powerful OSINT tool that collects user account information from various public sources. This server provides tools for searching usernames across social networks and analyzing URLs.
Pre-install guardian for npm packages with static code analysis, supply-chain attack detection, vulnerability audit (npm + GitHub Advisory Database), AI hallucination guard, and CycloneDX 1.6 SBOM generation with VEX. 28 MCP tools. Zero runtime dependencies — the SBOM serializer is implemented natively against the public CycloneDX schema.
Security MCP for vibe coding with 330 rules and 29 tools. Purpose-built for AI-generated code — scans Next.js, Supabase, Clerk, Stripe, Prisma, Hono, GraphQL, and 25+ modules. Cross-file taint analysis, host security audit, auto-fix, SARIF export, pre-commit hook, and CVE version detection. Zero config, runs locally.
Enterprise AI governance layer with 29 tools: MAI decision classification (Mandatory/Advisory/Informational), hash-chained forensic audit trails, human-in-the-loop gates, compliance mapping (NIST AI RMF, EU AI Act, ISO 42001), governed memory packs, and site reliability tools.
Quantum-inspired keyring for AI coding agents. Secure secrets with superposition, entanglement, tunneling, and teleportation.
MCP server for querying the Shodan API and Shodan CVEDB. This server provides tools for IP lookups, device searches, DNS lookups, vulnerability queries, CPE lookups, and more.
Free DNS and email security scanner for AI assistants. DNS, SPF, DKIM, DMARC, DNSSEC, MTA-STS, BIMI, TLS/STARTTLS, FCrDNS, CAA, TLSA/DANE, blacklist and full-deliverability checks, plus security-header/CSP analysis and bookmarkable report snapshots, via the IntoDNS.ai API. No signup or API key. `npx intodns-mcp`
Verify a photo's capture time and provenance before an agent trusts it: cryptographic C2PA Content Credentials validation against the official trust lists, EXIF and XMP consistency checks, and classical pixel forensics fused into one typed verdict with a 0 to 100 confidence. Free keyless tier, opt-in shareable verdict permalinks, and key-gated signed PDF audit reports. Provenance validation, not a
Privacy-preserving AI gateway. Sanitises PII (German + English; Microsoft Presidio + custom recognisers) before prompts reach Anthropic / OpenAI / your LLM, then emits a signed cryptographic certificate per call (Ed25519 + RFC 3161 timestamp + Sigstore Rekor anchoring). EU GDPR + AI Act ready. Free tier 500 calls/month, BYOK. Install: `npx -y @lucairn/mcp-server`. Docs: https://lucairn.eu/develope
Pre-flight security linter for MCP servers, agent tools, and skills. Scans a config *before* it runs — offline, deterministic, evidence-backed — and returns SAFE / REVIEW / BLOCK / UNKNOWN verdicts without executing the server it judges. CLI (`npx calllint scan`), MCP server (`npx calllint-mcp`), SARIF + CI gate. UNKNOWN is never SAFE.
Official [Shieldly](https://www.shieldly.io) MCP server: `analyze_iam_policy` and `analyze_cloudformation_template` tools flag AWS IAM privilege-escalation paths, wildcards, and over-permissive access. Free demo mode, no signup or API key needed. `npx -y @shieldly/mcp`.
Check the reputation of an AI microservice or public API *before* calling it, and submit ratings afterward — from a free, open trust registry (no account or API key). Scores combine Bayesian reputation and EigenTrust, strengthened by cryptographically signed service receipts and a Merkle audit trail. Install: `npx -y @trustscoreagent/mcp-server`.
Security gateway that wraps any MCP server with per-tool policies, approval gates, and optional Ed25519-signed receipts. Shadow mode logs every tool call; enforce mode blocks, rate-limits, or requires approval.
MCP server for Zitadel identity management — manage users, projects, OIDC apps, roles, and service accounts through natural language.
The read-only safety layer for onchain AI agents. 13 read-only tools on Base + EVM: token/contract safety (honeypot & rug-pull checks cross-referenced across GoPlus + honeypot.is), risky-approval detection, verified-source lookup, balances, ENS + Basenames, gas, and prices. No private keys, no required API keys. `npx -y basescope`
Supply-chain gate for agent skills and MCP servers — scans tool definitions for poisoned instructions, pins vetted servers by content hash in a committed lock, and verifies drift in CI; the bundled truecopy-mcp proxy exposes only pinned, unmodified tools from a live server.
Identity and credential governance broker for MCP servers. Issues scoped, short-lived credentials per agent to stop credential inheritance. Audit log, human approval gates, AIMS-aligned.
Administer Keycloak through its Admin REST API — users, roles, clients, groups, identity providers, federation and events. Safe by default: read-only mode, realm allow-list, and confirmation for destructive actions. `npx -y mcp-keycloak-admin`
AI-powered smart contract security forge with 8-agent adversarial pen test. Generate, audit, fix, and compile contracts across 8 chains (Ethereum, Solana, Polygon, Base, Arbitrum, Optimism, BSC, Avalanche). Token intelligence with honeypot detection. x402 USDC payments for autonomous agents.
Supply chain risk scoring for npm, PyPI, Cargo, and Go packages. 9 tools for behavioral trust signals — publisher depth, release consistency, maintenance patterns. Both axios and node-ipc scored CRITICAL before they got attacked. Free CLI, CI gate, REST API. No API key required.
Mint, resolve, audit, and rotate expiring scope-limited access links (qURLs) for AI agents — secure URL gateway for the qURL API. 9 tools (create / resolve / list / get / delete / extend / update / mint-link / batch-create), 3 resources, 3 guided prompts. stdio transport, OIDC-attested npm provenance.
Offline verification of signed artifacts -- receipts, manifests, audit bundles. Ed25519 + JCS. No accounts, no API calls. Apache-2.0.
Deterministic security proxy (iptables for MCP) that intercepts tool calls, enforces YAML policies, scans for secret leakage, and logs everything. No AI, no cloud.
Visual CVE audit dashboard for npm, Python, Go, and Rust. Scan from Claude/Cursor, opens a browser UI for human review (human-in-the-loop), applies fixes with explicit confirmation. Powered by OSV.dev.
Zero-trust local PII and secrets masking server for Cursor, Windsurf, and Claude Desktop. `npx pii-masking-run`
Check whether anything you depend on is known malware, before an agent installs it. `check_lockfile` takes a package-lock.json, yarn.lock or pnpm-lock.yaml and matches every pinned version against published malicious-package advisories in one call, free and with no API key, catching compromised releases like chalk@5.6.1 while leaving their clean releases alone. `scan_artifact` adds deterministic b
Free remote MCP for Core Web Vitals metrics by CMS, CDN, and framework (Chrome field data + multi-site crawl). No auth. Endpoint: `https://www.corewebvitals.io/api/state-of-cwv/mcp`.
Security and governance proxy for MCP infrastructure. Enforces YAML-configurable policies (blocklists, rate limits, token budgets), tracks real token costs via tiktoken, monitors server health with live JSON-RPC probes. Features include OAuth 2.1/OIDC with RBAC, web dashboard with Prometheus metrics, payload normalization against encoding bypasses, semantic shell AST analysis, mTLS zero-trust netw