hemlock

M4NT/hemlock

★ 1 stars Python AI/LLM Updated today

Complete security framework for AI Agents, MCP Servers, and RAG pipelines. Features automated MCP fuzzing, cross-agent poisoning detection, persistent memory audits, unified risk scoring, and CI/CD security gates.

View on GitHub → 🔍 Audit Wallet Slippage →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "hemlock": {
      "command": "uvx",
      "args": [
        "hemlock"
      ]
    }
  }
}

Or install with pip: pip install hemlock

README Excerpt

**Adversarial evaluation framework for LLM/RAG pipelines.** Hemlock is a Python research lab for measuring and defending against document-level attacks on Retrieval-Augmented Generation (RAG) systems. It provides: - **27 attack modules** across 8 categories (prompt injection, citation forgery, jailbreak-via-context, temporal spoofing, cross-tenant poisoning, semantic backdoor, memory/tool/graph attacks, and more)

Tools (7)

chromadbcompositefulllegacyregexsemanticsemantic_backdoor

Topics

ci-gatecross-agentmcpmemoryrag

hemlock

Quick Install

README Excerpt

Tools (7)

Topics

Related AI/LLM Servers