Security MCP servers

MCP policy proxy: enforces a code of conduct across connectors before the upstream is called.

Read-only Postgres/Supabase RLS audit with a rolled-back probe that proves who can read which rows.

Pre-execution security for AI agents on Base: guard token/tx/address + per-agent spend firewall.

Token-efficient Ghidra RE: decompile, xrefs, Swift/ObjC, ELF/Mach-O, async analysis

75-tool MCP server for 12 Kinetic Gain specs and DefenseTech checks.

Live cross-domain authority-state resolution for delegated agents through MCP.

Local privacy-first tools: redact secrets, decode JWTs, compare LLM costs. No network calls.

Keep patient identifiers out of AI agent context — a local privacy boundary for healthcare agents.

MCP server: MITRE ATT&CK® -> NIST 800-53r5 -> DISA STIG fix/check steps, mitigations and detections

Agent 运行时全栈拦截: allow/deny tool calls, Ed25519 receipts, blocks RCE, theft, SSRF, injection. stdio.

Check an MCP server before you install it — outside-in trust ratings for 7,000+ operators.

Read and change VPN Bypass (macOS) routing: domain lists, services, modes, kernel routes, logs

Community-built, comprehensive MCP server for the SecurityScorecard API (unofficial).

Policy-checked, audited agent access to Postgres, MySQL, SQLite, MongoDB and OpenAPI REST APIs.

Finds lethal-trifecta toxic flows in MCP tool catalogs: untrusted input, sensitive data, egress

Offline full-text search over the HackTricks security wiki, synced every 3 days.

Intended's authority layer as MCP tools: ALLOW/ESCALATE/DENY decisions with signed tokens.

Injects secrets into shell commands at exec time; plaintext never reaches the LLM context

Sourcery security findings: MCP server, CLI, and fix-prompt builder over the Sourcery API.

Centerfield building (Seoul) visitor pre-registration: validate, preview and register visitors.

Sourcery security findings: MCP server, CLI, and fix-prompt builder over the Sourcery API.

DEPRECATED: no live endpoint (api.unico.co does not resolve)

Sits in front of any MCP server. Reads pass, writes are checked against your lists.

AI-safe credential broker for MCP — run saved SSH/MySQL/psql aliases; passwords never reach AI.

Tamper-evident memory + governance MCP for Claude Code, Cursor, Codex. Ships the Seatbelt kit.

Security sidecar for MCP servers: 11 read-only control-plane tools for prompt-injection checks.

Precision-first security scanning for Model Context Protocol servers.

MCP server for CTFd v3: list challenges, submit flags, and query scoreboards via AI agents.