- Conduid
- Marketplace
- #ai-safety
MCP servers tagged ai-safety
464 live MCP servers tagged "ai-safety", ranked by trust score. Tags come from package metadata and repository topics, so this list covers servers across every category that work with ai-safety.
Promptwall
A firewall for LLM traffic: detect prompt injection, jailbreaks, and secret/PII leaks. Zero-dependency Python library + CLI + MCP server.
AI Tutor
MCP server for AI Tutor Card disclosures. Six tools for procurement review, curriculum matching, and FERPA / COPPA compliance auditing of AI tutors. EdTech-flavored extension of t…
AI Skill MCP Audit Protocol
A safe review protocol for external AI skills, MCP servers, and prompt packages.
Vexa Agent Control
Vexa Agent Control is an open source AI Gateway and transparent security sidecar purpose-built for AI agents, developers, and enterprise platform teams. It operates in two flexibl…
Axiom Guardian
AI Guardian that challenges actions before execution. Forces clarity of thought through adversarial questioning and NLI-based reasoning evaluation.
Wordpress MCP Agent Bridge
WordPress MCP + REST bridge for AI agents — verified Rank Math SEO and schema writes, hashed snapshots, verified restore, additive-only gates. Claude Code, Cursor, any MCP client.…
Glassport
Wire-level observability and enforcement for MCP servers: passive stdio tap, behavioral detectors, drift watch, and static audit — with SARIF for the GitHub Security tab. Zero dep…
Grounding Enforcer
A standalone MCP suppressor that enforces grounding by validating references, rejecting unsupported claims, and stabilising model outputs.
Psychopathia MCP
MCP server exposing the Psychopathia Machinalis diagnostic framework (79 conditions, 11 tools) so AI systems can diagnose dysfunctions in themselves or others. PyPI: psychopathia-…
Csoai Watermark Attest MCP
💧 Exposes EU AI Act Article 50 watermarking and disclosure checks over MCP — C2PA, content provenance, deepfake disclosure, AI-content labelling. MIT.
Shield Pii
🛡️ Real-time PII redaction proxy for MCP (Model Context Protocol) — detects and masks 23 entity types before they reach the LLM. Drop-in privacy layer for Claude Desktop with zero…
Wild Rails AI Ops
Stop hand-rolling agent-safety code in every Rails app. Wild gives Rails teams 10 production-ready Ruby gems for safe AI-agent ops: capability gating, audited execution, MCP serve…
Orgai Platform
Self-hosted AI compliance enforcement — one policy source enforced across Claude Code, Cursor, Copilot & any MCP agent, with git-hook + CI gates and an append-only audit trail
Policy Gateway
Runtime policy enforcement and adversarial evaluation for MCP tool calls
Nervous System Action
GitHub Action for Nervous System LLM behavioral enforcement - checks PRs against governance rules
Pfc Execution Boundary Proof
Public proof surface for PFC-style execution-boundary governance and signed runtime admission receipts.
Safe MCP Suite
Two hardened MCP servers - a terminal executor and a file organizer - on one shared deny-by-default safety core, with a fail-closed audit trail and undoable file moves.
Caucus
Auditable multi-agent consensus grounded in live MCP evidence, with recorded dissent and tamper-evident decision logs.
AI Governance
MCP server exposing AI governance as callable tools. Control mapping across NIST AI RMF, ISO/IEC 42001, and the EU AI Act, readiness scoring, risk-tier classification, and remedia…
Belay MCP
Transaction firewall for AI agents. Preview, govern, verify and rewind MCP and native tool calls with deterministic policies, human approvals and signed evidence.
Evidence First Agents
Deterministic readiness and boundary checks for AI-agent projects across skills, MCP, tools, approvals, and handoffs.
Contextlock
Local MCP safety layer that blocks secrets before AI coding agents see them.
Privacy Shield MCP
7-layer security plugin for Claude Code: real-time hooks (bash, file, web) + 21 MCP tools for privacy, threat detection, and audit
Nornr Sentry
NORNR Sentry local airbag for dangerous agent actions
Tex
Runtime governance for AI agents — PERMIT / ABSTAIN / FORBID enforced in-path, sealed with tamper-evident receipts that verify offline.
AI Companion Chatbot Demo
Voice-first conversational AI demo with layered prompts, safety guardrails, MCP tool-calling, and a from-scratch eval framework.
Two Tier Safe AI Gate
Two-tier safe execution model for LLM agents. Go + Inngest + Omnigent MCP. Policy-gated durable workflows.
Tulip Agents
The agent framework where the model never holds the trigger — every consequential action clears your policy first, waits for a human when it matters, and lands on a record you can…
Presshub
Enterprise AI Co-Pilot & Multi-Provider LLM Orchestration Engine with Automated 0–100 Readiness Gates, Sliding-Window Rate Limiting, and Deep-Research.
Spooky Action
Computer use on a Mac for any AI agent. It presses the button, not a pixel — through the accessibility layer, so your cursor never moves. Watch it work, and take the screen back b…
epistemedeus/skillguard
y github:epistemedeus/skillguard mcp`
Promptmirror
Prompt-injection & indirect-injection scanner for any LLM context input
Hardened Terminal MCP
Security-hardened MCP server for terminal access: deny-by-default policy engine, shell=False sandboxed execution, cwd jail, secret redaction, fail-closed audit log.
Pandaone AI Agent
Pandaone Guard — Free open-source MCP server for AI code audit. L1-L6 defense (file lock + audit log + pre-commit hooks) for Claude/Cursor/Trae. Local stdio, no API key, MIT.
Boundary Audit
Detects MCP tools hidden from tools/list yet reachable via tools/call; the bundled mock reproduces FAIL and PASS. Actively calls operator-named tools; 11 tests, CI. Does not disco…
Hallucination Suite
A unified suite of MCP suppressors that prevent hallucinations, enforce grounding, and stabilise agent behaviour.
Depscope CLI
CLI for depscope.dev — audit deps before install
Antidote
Security scanner for MCP tool servers that detects poisoning, injection surfaces, and propagation paths
Protect MCP Adk
Ed25519 receipt signing plugin for Google ADK agents. BasePlugin with after_tool_callback for cryptographic decision receipts.
Veriethiccore MCP
VeriEthicCore — EU AI Act + Halal Compliance MCP Server. The world's first MCP combining EU AI Act risk scanning with Islamic Sharia ethics validation.
SMERC Runtime Permission Layer
Pre-execution recoverability control for AI agents, MCP tool calls, cloud automation, and high-impact workflows.
Agent Eval Gate
MCP eval-gate server: an agent's only path to moving money runs a fail-closed, version-pinned OFAC screen-before-transfer control first. Stdlib-only, zero-install, polarity-proven.
Magneto
Magneto is a non-LLM adversarial citation gate whose job is to prevent hallucinations by fact-checking an agent in real-time. Currently for Kiro. EXPERIMENTAL.
Memoryos Guardian Core
Experimental Safe Execution Core for AI agents — state-bound execution and evidence-based outcomes.
Webmcp Sentinel
Server-enforced authorization, human approval, and audit for AI-agent actions — a WebMCP security reference implementation
Sqlgate
A deterministic, fail-closed write gate for LLM-generated SQL, exposed over MCP. The model can't talk its way past it.
Steerpack
Rule lifecycle governance for LLM agents — human adjudication pipeline, mid-execution watchdog, cross-harness portable export.
Lens Kit
Trainable 10-lens validation gate for AI-generated text. Outside-in: a separate model checks the output, fixed checks, inspectable findings, fail-closed. Apache-2.0 — own it and t…
Cid
L.ai · Gate (CID) — per-token validation for LLMs. Catches hallucinations, math errors, and fallacies in real time. 5+ gates, 1,606 facts, 630KB WASM.
AI Policy Platform
Self-hosted AI compliance enforcement — one policy source enforced across Claude Code, Cursor, Copilot & any MCP agent, with git-hook + CI gates and an append-only audit trail