Agentshield Mcp

Overview

Agentshield Mcp is a Model Context Protocol server that detects prompt injection, jailbreak, and social-engineering attacks targeting LLM agents. When installed, any MCP-compatible AI assistant can invoke its security tools directly, letting the model decide when to apply protections during conversations or agent runs.

The server runs locally as an npm package. Install it with npx -y @eigenart/agentshield-mcp and configure it using the AGENTSHIELD_API_KEY environment variable. Version v0.1.3 is available under the MIT license, with source code hosted at dl-eigenart/agentshield-platform on GitHub.

Agentshield Mcp appears in the AI & Agents category and integrates with all major MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, and Zed. Maintainers from dl-eigenart are welcome to claim this listing to verify ownership and keep information up to date.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "agentshield_mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@eigenart/agentshield-mcp"
      ],
      "env": {
        "AGENTSHIELD_API_KEY": "<agentshield-api-key>"
      }
    }
  }
}

Requires AGENTSHIELD_API_KEY