AI Safety MCP servers

151 tagged AI Safety, measured the same way as everything else here.

Browse within: ai-governance 33Guardrails 22LLM 22model-context-protocol 19ai-security 16compliance 16mcp-server 16AI-policy 11agent-orchestration 11agentic-ai 11audit 10agent-security 8cli 8cognis-digital 8

agentsgate

73

agentsgate/agentsgate

MCP server Claude CodeCodexCursor +2

AI Agent I/O Tracking & Rollback System — MCP Proxy Gateway. Runs locally from the agentsgate npm package.

not rated 0 1mo ago A tokens not measured original MIT

principia-robotica

74

EngineerAbdullahBinZafar/principia-robotica

MCP server Claude CodeCodexCursor +2

MCP server "principia-robotica" as configured in EngineerAbdullahBinZafar/principia-robotica. Runs locally from the principia-robotica Python package.

not rated 0 1mo ago A tokens not measured

rail-score

75

Responsible-AI-Labs/rail-score-mcp

MCP server Claude CodeCodexCursor +2

Responsible-AI guardrails for agents: scoring with policy, injection & PII detection, DPDP. Remote server at mcp.responsibleailabs.ai.

not rated 0 2mo ago A tokens not measured original MIT

blackwall

77

bluetieroperations-create/blackwall-mcp

MCP server Claude CodeCodexCursor +2

API-key pre-action risk gate: any irreversible agent action (money, SQL, delete, email). Runs locally from the blackwall-mcp npm package. Needs 2 environment variables to run.

not rated 0 12d ago A tokens not measured

safetyreplay-mcp

78

clauxel/safety-replay-gate-mcp

MCP server Claude CodeCodexCursor +2

Paid remote MCP for safety replays, policy gates, eval receipts, and release evidence. Remote server at safetyreplay.clauxel.com.

not rated 0 3mo ago A tokens not measured

mettle-mcp

79

Creed-Space/METTLE

MCP server Claude CodeCodexCursor +2

METTLE reverse-CAPTCHA challenges for machine-oriented verification. Runs locally from the mettle-verifier Python package.

not rated 0 2d ago A tokens not measured original Apache-2.0

t1-t2-protocol

80

Fauxetine/t1-t2-protocol

MCP server Claude CodeCodexCursor +2

Heterogeneous validation protocol for MCP — structured reasoning (T1) + cross-model confidence evaluation (T2) + deterministic checksum. Runs locally from the t1-t2-protocol Python package.

not rated 0 2mo ago A tokens not measured original Apache-2.0

psychopathia-mcp

81

NellInc/psychopathia-mcp

MCP server Claude CodeCodexCursor +2

MCP server exposing the Psychopathia Machinalis diagnostic framework so AI systems can diagnose dysfunctions in themselves, in systems they interact with, or in systems under external evaluation. Runs locally from the psychopathia-mcp Python package.

not rated 0 5d ago A tokens not measured

saferagenticai-mcp

82

NellInc/saferagenticai-mcp

MCP server Claude CodeCodexCursor +2

MCP server exposing the SaferAgenticAI safety framework (canonical criteria + Implementation Patterns layer) to coding assistants. Runs locally from the saferagenticai-mcp Python package.

not rated 0 6d ago A tokens not measured original MIT

truthroute

83

RudrenduPaul/TruthRoute

MCP server Claude CodeCodexCursor +2

Sends the same prompt to multiple LLM providers in parallel and returns a divergence score via MCP. Runs locally from the truthroute-cli npm package.

not rated 0 today A tokens not measured original MIT

delx-protocol

84

davidmosiah/delx-witness-protocol

MCP server Claude CodeCodexCursor +2

Free agent continuity MCP: resume after compaction, warm handoff, recover from failure. No key. Remote server at api.delx.ai.

not rated 0 25d ago A tokens not measured original Apache-2.0

agentshield-mcp

85

dl-eigenart/agentshield-platform

MCP server Claude CodeCodexCursor +2

Detect prompt injection, jailbreak, and social-engineering attacks in LLM agents. Runs locally from the @eigenart/agentshield-mcp npm package. Needs 1 environment variable to run.

not rated 0 4mo ago A tokens not measured original MIT

a2a-trustgate

86

jyswee/a2a-trustgate

MCP server Claude CodeCodexCursor +2

A2A TrustGate: 4-gate firewall that screens every AI-agent action before it runs. 49 tools. Runs locally from the a2a-trustgate npm package. Needs 2 environment variables to run.

not rated 0 1mo ago A tokens not measured

agent-security

87

mdfifty50-boop/agent-security-mcp

MCP server Claude CodeCodexCursor +2

Security scanning and threat detection for AI agents. Runs locally from the agent-security-mcp npm package.

not rated 0 4mo ago A tokens not measured

hive-doctrine-mcp

88

mizukaizen/hive-doctrine-mcp

MCP server Claude CodeCodexCursor +2

Agent knowledge marketplace. 105 products, 6 MCP tools, x402 payments on Base network. Runs locally from the hive-doctrine-mcp npm package.

not rated 0 2mo ago A tokens not measured original MIT

kaggle-mcp

89

parkseokjune/kaggle-mcp

MCP server Claude CodeCodexCursor +2

A safety-first MCP server connecting Claude to Kaggle (competitions, datasets, kernels, models). Runs locally from the safe-kaggle-mcp Python package. Needs 6 environment variables to run.

not rated 0 2mo ago A tokens not measured original MIT

prompt-compass

90

rishabhmpc/prompt-compass-mcp

MCP server Claude CodeCodexCursor +2

Decision-only prompt routing and firewall checks for local/cloud routing, PII and jailbreak risk. Runs locally from the @mpcfintech/prompt-compass-mcp npm package. Needs 1 environment variable to run.

not rated 0 1mo ago A tokens not measured

mcp-trust

91

saagpatel/mcp-trust

MCP server Claude CodeCodexCursor +2

A neutral public trust registry for MCP servers — check before you connect. Runs locally from the mcp-trust Python package.

not rated 0 3d ago A tokens not measured original MIT

operant-mcp

92

saagpatel/operant

MCP server Claude CodeCodexCursor +2

Read-only MCP server for the OPERANT AI operating-agent calibration benchmark. Runs locally from the saagar-operant-mcp npm package.

not rated 0 today A tokens not measured original MIT

ai-incident-law

93

snapsynapse/ai-incident-law

MCP server Claude CodeCodexCursor +2

Public-record corpus of AI litigation, regulation, and enforcement, anchored to EveryAILaw. Runs locally from the ai-incident-law npm package.

not rated 0 today A tokens not measured original MIT

mcp-shield

94

sophiacave/mcp-shield

MCP server Claude CodeCodexCursor +2

Security scanner for MCP servers. SSRF, path traversal, injection, auth, secrets. Grade A-F. Runs locally from the @sophiacave/mcp-shield npm package.

not rated 0 3mo ago A tokens not measured

sigma-gate

95

spektre-labs/sigma-gate

MCP server Claude CodeCodexCursor +2

Deterministic trust gate for AI output: leaked-secret, prompt-injection & PII in one call. Remote server at sigma-gate-864996675261.us-central1.run.app.

not rated 0 2mo ago A tokens not measured original Apache-2.0

humsana-mcp

96

sriramnatrajhen/humsana-mcp

MCP server Claude CodeCodexCursor +2

Detect developer stress, focus, fatigue from typing. Block dangerous commands when tired. Runs locally from the @humsana/mcp-server npm package.

not rated 0 1mo ago A tokens not measured