Skip to content
Agentmart
Submit a tool

Best MCP servers for guardrails and AI safety

We found 100 MCP servers for guardrails and AI safety in the catalog; these are the 15 that rank highest. The order comes from our rank, which weighs adoption, freshness, documentation and publisher identity. Nobody can pay for a place on this list. MCP Hangar, GroundLens and Midplane lead it. MCP Hangar, Midplane and Safeprompt are published by the company that makes them. MCP Hangar has the most GitHub stars here (18). 14 of the 15 had activity in the last 30 days. Remote servers need no local install; local ones run on your machine and can reach local files. Each entry lists why it made the list and links to its install page.

Updated Oct 26, 2026 · Built from catalog data; how we rank and grade

Top 15 of 100

  1. 1
    MCP Hangar

    mcp-hangar.ioOfficial

    Policy enforcement plane for MCP: every tool call ends in a verdict.

    • Published by the vendor
    • Most GitHub stars here (18)
    • Most weekly downloads here (1.3K)

    ★ 18↓ 1.3K/wkLast active Oct 25, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  2. 2
    GroundLens

    groundlens-dev

    Guardrails for LLM answers: exact checks against sources, PASS/REVIEW/FAIL, signed records.

    • Active in the last 30 days
    • Open source (Apache-2.0)

    ★ 7↓ 318/wkLast active Oct 19, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  3. 3
    Midplane

    midplane.aiOfficial

    Safe-by-default SQL guardrails for AI agents: AST-checked queries, per-table policy, audit log.

    • Published by the vendor
    • Runs locally or as a remote server
    • Active in the last 30 days

    ★ 5Last active Oct 3, 2026Local (stdio), Remote (HTTP)Claude Code, Codex, VS Code, Cursor +2

  4. 4
    LLM Redteam

    GuardBee

    Active jailbreak/extraction/obfuscation red-teaming for live LLM endpoints

    • Active in the last 30 days
    • Open source (MIT)

    ★ 1↓ 618/wkLast active Oct 22, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  5. 5
    Cost Guard MCP

    mcpsmiths

    Pre-flight query cost and result-size guardrails for AI agents on BigQuery, Snowflake, Databricks

    • Active in the last 30 days
    • Open source (MIT)

    ★ 1↓ 216/wkLast active Oct 19, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  6. 6
    Superuser

    fitzyracing1

    Permission-gated superuser (sysadmin) MCP agent with enterprise guardrails and audit.

    • Active in the last 30 days
    • Open source (Apache-2.0)

    ★ 3↓ 125/wkLast active Oct 22, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  7. 7
    Safeprompt

    safeprompt.devOfficial

    Detect prompt injection, jailbreaks, and code injection in untrusted text before it reaches an LLM.

    • Published by the vendor
    • Active in the last 30 days
    • Open source (MIT)

    ★ 2Last active Oct 6, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  8. 8
    Diogenes Autonomous Microservice Gateway

    stand-out-of-my-sun

    Autonomous AI microservices: 55 dynamic guardrails, contract audits & oracles via L402/x402/MPP.

    • Remote server, no local install
    • Active in the last 30 days
    • Open source (MIT)

    ★ 2Last active Oct 23, 2026Remote (HTTP)Claude Code, VS Code, Cursor, Windsurf

  9. 9
    Promptrejectormcp

    revsmoke

    Security gateway for AI agents: detects prompt injections, jailbreaks, and common vulnerabilities.

    • Active in the last 30 days
    • Open source (ISC)

    ★ 2Last active Oct 1, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  10. 10
    Agentrisk

    trycoin-ai

    Risk guardrails for AI trading agents.

    • Active in the last 30 days
    • Open source (MIT)

    ★ 2↓ 56/wkLast active Sep 28, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  11. 11
    Stratum MCP

    smartmemory.ai

    Structured execution for coding agents: contracts, postconditions, gates, guardrails.

    • Active in the last 30 days

    ★ 1Last active Oct 24, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  12. 12
    Koma Gate MCP

    swnotmetal

    Classifies prompt injection, jailbreaks, and out-of-scope user input before agents act on it.

    • Open source (MIT)

    ★ 13Last active Aug 18, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2

  13. 13
    signet-eval

    jmcentire

    Deterministic policy enforcement and MCP management for AI agent tool calls.

    • Active in the last 30 days
    • Open source (MIT)

    ★ 4Last active Oct 13, 2026

  14. 14
    DCL Trust Oracle — AI/LLM Output Audit (x402 MCP)

    fronesislabs.com

    AI/LLM agent output audit MCP: policy eval, tamper-evident chain, AI safety, x402 USDC on Base.

    • Remote server, no local install
    • Active in the last 30 days
    • Open source (Apache-2.0)

    ★ 1Last active Oct 17, 2026Remote (HTTP)Claude Code, VS Code, Cursor, Windsurf

  15. 15
    WhitePact

    Guruprasath-Annadurai

    AI governance MCP server: trust scoring, guardrails, bias/hallucination detection, compliance.

    • Runs locally or as a remote server
    • Active in the last 30 days
    • Open source (MIT)

    ★ 2↓ 37/wkLast active Oct 11, 2026Local (stdio), Remote (HTTP), Remote (SSE)Claude Code, VS Code, Codex, Cursor +2

Related