Security middleware for AI agent frameworks. Scans packages against CVE databases, malicious package lists, license policies, and trust scores before installation. Supports pip, npm, and cargo. Integrates via MCP server, CLI, pre-commit hook, GitHub Action, HTTP daemon, IPC socket, or interactive shell guard. Fully local, no telemetry.

Quick Install

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

Add to claude_desktop_config.json:

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

Or install with pip: pip install agentshield

README Excerpt

**Security layer for AI agent frameworks.** AgentShield intercepts package installation requests made by AI agents, checks them against CVE databases and static analysis tools, enforces configurable response policies, and generates security posture reports — all locally, with no telemetry. > **AI agent?** Skip straight to the **[Agent Setup Guide](AGENT_SETUP.md)** — it has everything your agent needs to integrate AgentShield in one self-contained document.

Tools (20)

AGENTSHIELD_OFFLINEAGENTSHIELD_SESSION_IDALLOWBLOCKGETGITHUB_TOKENLOG_ASYNCNEEDS_CONFIRMATIONNVD_API_KEYPOSTagentshield_diff_scanagentshield_postureagentshield_sbomagentshield_scanagentshield_scan_dockeragentshield_scan_fileallowlistasync_reportbandit_runnerblock

AgentShield

Quick Install

README Excerpt

Tools (20)

Related Code/Dev Tools Servers