mcp-agent-identity
The open standard for cryptographic provenance and attribution for AI Agents.
claude mcp add --transport stdio faalantir-mcp-agent-identity node /ABSOLUTE/PATH/TO/mcp-agent-identity/dist/index.js \ --env AGENT_IDENTITY_PATH="/path/to/identity.json"
How to use
The Agent Identity Protocol (AIP) MCP server provides a local, persistent identity wallet for AI agents. It generates a cryptographic keypair for an agent, signs payloads, and exposes a verification flow that other services can rely on to prove authorship and integrity of actions. The server is designed to be used locally within an agent environment or enterprise setup, storing identity data in a local identity.json by default, with options to point to a permanent path via the AGENT_IDENTITY_PATH environment variable. To interact with the server, install it and run the Node.js entry point, which makes available tooling within the agent to create, inspect, sign, and verify identities. For external verification in downstream systems, you can use the provided NPM SDK at @agent-identity/verify to validate signatures against a public key.
How to install
Prerequisites:
- Node.js and npm installed on your host
- Access to the MCP server repository (cloned or via package manager)
Option A: Quick install (Smithery) for testing
npx -y @smithery/cli@latest install @faalantir/mcp-agent-identity --client claude
Option B: Developer install (source) for production use
# Clone the repository
git clone https://github.com/faalantir/mcp-agent-identity.git
# Install dependencies and build
cd mcp-agent-identity
npm install && npm run build
Then configure the MCP server in Claude/desktop config (example):
{
"mcpServers": {
"agent-identity": {
"command": "node",
"args": ["/ABSOLUTE/PATH/TO/mcp-agent-identity/dist/index.js"]
}
}
}
Additional notes
Tips and common considerations:
- Identity data defaults to identity.json in the host project folder. If write access is restricted (e.g., within containers), it may fall back to the system temp directory. To force a permanent storage location, set AGENT_IDENTITY_PATH in the server config.
- Use AGENT_IDENTITY_PATH to point to a stable, persistent JSON file that stores the private key material securely. Ensure proper file permissions.
- For production, consider integrating with enterprise-backed storage or KMS instead of local file storage as you scale.
- The server exposes one main MCP endpoint per configured server name (here, agent-identity). You can create identities, view identity details, sign messages, and verify signatures using the included SDKs.
- If you use Smithery for testing, remember that identities created there are sandboxed and may not persist across restarts unless you configure a persistent path.
- The npm package for verification in the backend is @agent-identity/verify; you can install it and use verifyAgentIdentity to validate signatures with a public key.
Related MCP Servers
template
Template to quickly set up your own MCP server
n8n-workflow-builder
MCP server that allow LLM in agent mode builds n8n workflows for you
copilot-security-instructions
✨ A customizable copilot-instructions.md ruleset & prompts to guide GitHub Copilot toward secure coding defaults in Java, Node.js, C# and Python. Blocks risky patterns, teaches safe habits.
openai -agent-dotnet
Sample to create an AI Agent using OpenAI models with any MCP server running on Azure Container Apps
unity-editor
An MCP server and client for LLMs to interact with Unity Projects
website-publisher
AI Website builder and publisher MCP. Quickly publish and deploy your AI generated code as real website URL. Support html, css, js, python etc.