agentguard

doctormizio777777/agentguard

★ 0 stars Python AI/LLM Updated today

The intelligent firewall for AI agents — a GPT-5.6 intent layer, tamper-evident audit ledger, and MCP server that catch hijacked agents static rules can't see.

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": {
    "agentguard": {
      "command": "uvx",
      "args": [
        "agentguard"
      ]
    }
  }
}

Or install with pip: pip install agentguard

README Excerpt

**Live demo — start here:** [AgentGuard landing](https://agentguard-dusky.vercel.app) · [Mission Control console](https://agentguard-dusky.vercel.app/console) · [Run locally with Docker](#3-try-it-in-60-seconds) AI agents act autonomously across payments, email, data, APIs, and system tools. Prompt injection can hijack that autonomy and make a legitimate agent serve an attacker. Static spend rules can cap money, but they cannot see whether the agent's intent still matches its mission. Gartner pr

agentguard

Quick Install

README Excerpt

Related AI/LLM Servers