Sponsored by Deepsite.site

Identity

Created By
agntcy6 months ago
AGNTCY Identity allows to onboard, create and verify identities for Agents, Model Context Protocol (MCP) Servers and Multi-Agent Systems (MASs).
Content

Identity

Lint Contributor-Covenant


Welcome to the Identity repository


AGNTCY Identity provides a secure and verifiable method to uniquely identify agents through open and decentralized techniques. Each agent is assigned a universally unique identifier, backed by verifiable credentials (VCs). AGNTCY Identity enables to bring your own identity using conventions like IDs assigned by Identity Providers (e.g., Okta) or Agent Cards (e.g., Google’s A2A), or be assigned an ID following standards (e.g., W3C DIDs). This component ensures that every agent in the AGNTCY ecosystem has a verifiable, universally unique identity, enabling secure authentication, trusted communication, and interoperability across diverse multi-agent systems, regardless of the identity assignment method.

  • The ID is linked to a ResolverMetadata object for secure and automated verification.
  • The ID can be linked to one or more Agent Badges. Why? Multiple badges can provide nuanced, task-specific access to different systems without over-privileging the agent. Agent Badges contain Verifiable Credentials (VCs), which include:
    • The Agent's ID
    • Schema definition (e.g., OASF)
    • Metadata for authentication and other security needs.

NOTE

This same structure applies to MCP Servers and MASs, ensuring consistency across all identity-bearing entities in the IoA.

📚 Table of Contents

You can also:

🚀 Architecting Agentic Trust

  • Core Principle: Trust is foundational for the Internet of Agents.
  • Identity as the Root: AGNTCY Identity ensures Agents and Tools (MCP Servers) are verifiably authentic.
  • Flexible & Interoperable: BYOID (Bring Your Own ID), integrates with existing Identity Providers (IdPs).

Secure and reliable communication between software agents is a cornerstone of the Internet of Agents (IoA) vision. Without proper identity management, malicious or unverified agents can infiltrate Multi-Agent Systems (MASs), leading to misinformation, fraud, or security breaches. To mitigate these risks, the AGNTCY provides a standardized and consistent framework for authenticating agents and validating associated metadata. This applies equally to:

  • Agents
  • Model Context Protocol (MCP) Servers
  • MASs (Multi-Agent Systems)

TIP

This repository includes an AI Agent and MCP Server to showcase the AGNTCY Identity components in action!

🌟 Features & Main Components

Features

  • Identity creation: Generate unique, verifiable identities for agents and MCP servers.
  • Existing identity onboarding: Integrate identities from external IdPs.
  • Badges creation & verification: Authenticate agents and MCP servers and validate metadata.

Main Components

  • Issuer CLI: Manage identities, vaults and credentials via command-line interface.
  • Node Backend: Backend server for identity management and metadata.

⚡️ Get Started in 5 Minutes

This short guide allows you to setup the Identity Issuer CLI as well as the Identity Node Backend. The Issuer CLI allows to generate, register, search for, and verify badges for Agents and MCP Servers. The CLI includes a library enabling storage and retrieval of the keys required to sign the badges, both on local storage or using a 3rd party wallet or vault. The Node Backend comprises the APIs and the backend core. It stores, maintains, and binds org:sub-org IDs, PubKeys, Subject IDs and metadata, including badges, ResolverMetadata and Verifiable Credentials (VCs).

Prerequisites

To run these steps successfully, you need to have the following installed:

Step 1: Install the Issuer CLI

Use the following command to install the Issuer CLI:

using curl:

sh -c "$(curl -sSL https://raw.githubusercontent.com/agntcy/identity/refs/heads/main/deployments/scripts/identity/install_issuer.sh)"

or using wget:

sh -c "$(wget -qO- https://raw.githubusercontent.com/agntcy/identity/refs/heads/main/deployments/scripts/identity/install_issuer.sh)"

NOTE

You can also download the Issuer CLI binary corresponding to your platform from the latest releases.

On some platforms you might need to add execution permissions and/or approve the binary in System Security Settings.

For easier use, consider moving the binary to your $PATH or to the /usr/local/bin folder.

If you have Golang set up locally, you could also use the go install command:

go install github.com/agntcy/identity/cmd/issuer@latest && \
  mv $(go env GOPATH)/bin/issuer $(go env GOPATH)/bin/identity

Step 2: Start the Node Backend with Docker

  1. Clone the repository and navigate to the identity directory:

    git clone https://github.com/agntcy/identity.git && cd identity
    
  2. Start the Node Backend with Docker:

    ./deployments/scripts/identity/launch_node.sh
    

    Or use make if available locally:

    make start_node
    

Step 3: Verify the Installation

You can verify the installation by running the command below to see the different commands available:

identity -h

📜 Core commands to use the CLI

Here are the core commands you can use with the CLI

  • vault: Manage cryptographic vaults and keys
  • issuer: Register and manage issuer configurations
  • metadata: Generate and manage metadata for identities
  • badge: Issue and publish badges for identities
  • verify: Verify identity badges
  • config: Display the current configuration context

🧪 Run the demo

This demo scenario will allow you to see how to use the AGNTCY Identity components can be used in a real environment. You will be able to perform the following:

  • Register as an Issuer
  • Generate metadata for an MCP Server
  • Issue and publish a badge for the MCP Server
  • Verify the published badge

Prerequisites

First, follow the steps in the Get Started in 5 minutes section above to install the Issuer CLI and run the Node Backend, and generate a local vault and keys.

To run this demo setup locally, you need to have the following installed:

Step 1: Run the Samples with Ollama and Docker

The agents in the samples rely on a local instance of the Llama 3.2 LLM to power the agent's capabilities. With Ollama installed, you can download and run the model (which is approximately 2GB, so ensure you have enough disk space) using the following command:

  1. Run the Llama 3.2 model:

    ollama run llama3.2
    
  2. From the root of the repository, navigate to the samples directory and run the following command to deploy the Currency Exchange A2A Agent leveraging the Currency Exchange MCP Server:

    cd samples && docker compose up -d
    
  3. [Optional] Test the samples using the provided test clients.

Step 2: Use the CLI to create a local Vault and generate keys

  1. Create a local vault to store generated cryptographic keys:

    identity vault connect file -f ~/.identity/vault.json -v "My Vault"
    
  2. Generate a new key pair and store it in the vault:

    identity vault key generate
    

Step 3: Register as an Issuer

For this demo we will use Okta as an IdP to create an application for the Issuer. To quickly create a trial account and application, we have provided a script to automate the process using the Okta CLI.

IMPORTANT

If you already have an Okta account, you can use the okta login command to log in to your existing organization.

If registering a new Okta developer account fails, proceed with manual trial signup and then use the okta login command, as instructed by the Okta CLI.

  1. Run the following command from the root repository to create a new Okta application:

    . ./demo/scripts/create_okta_app
    
  2. In the interactive prompt, choose the following options:

    > 4: Service (Machine-to-Machine), > 5: Other

  3. Register the Issuer using the Issuer CLI and the environment variables from the previous step:

    identity issuer register -o "My Organization" \
        -c "$OKTA_OAUTH2_CLIENT_ID" -s "$OKTA_OAUTH2_CLIENT_SECRET" -u "$OKTA_OAUTH2_ISSUER"
    

NOTE

You can now access the Issuer's Well-Known Public Key at http://localhost:4000/v1alpha1/issuer/{common_name}/.well-known/jwks.json, where {common_name} is the common name you provided during registration.

Step 4: Generate metadata for an MCP Server

Create a second application for the MCP Server metadata using Okta, similar to the previous step:

  1. Run the following command from the root repository to create a new Okta application:

    . ./demo/scripts/create_okta_app
    
  2. In the interactive prompt, choose the following options:

    > 4: Service (Machine-to-Machine), > 5: Other

  3. Generate metadata for the MCP Server using the Issuer CLI and the environment variables from the previous step:

    identity metadata generate -c "$OKTA_OAUTH2_CLIENT_ID" \
        -s "$OKTA_OAUTH2_CLIENT_SECRET" -u "$OKTA_OAUTH2_ISSUER"
    

NOTE

When successful, this command will print the metadata ID, which you will need in the next step to view published badges that are linked to this metadata.

Step 5: Issue and Publish a Badge for the MCP Server

  1. Issue a badge for the MCP Server:

    identity badge issue mcp -u http://localhost:9090 -n "My MCP Server"
    
  2. Publish the badge:

    identity badge publish
    

NOTE

You can now access the VCs as a Well-Known at http://localhost:4000/v1alpha1/vc/{metadata_id}/.well-known/vcs.json, where {metadata_id} is the metadata ID you generated in the previous step.

(Optional) Step 6: Verify a Published Badge

You can use the Issuer CLI to verify a published badge any published badge, not just those that you issued yourself. This allows others to verify the Agent and MCP badges you publish.

  1. Download the badge that you created in the previous step, replacing {metadata_id} with the metadata ID from step 4:

    curl -o vcs.json http://localhost:4000/v1alpha1/vc/{metadata_id}/.well-known/vcs.json
    
  2. Verify the badges using the Issuer CLI:

    identity verify -f vcs.json
    

Development

For more detailed development instructions please refer to the following sections:

Roadmap

See the open issues for a list of proposed features (and known issues).

Contributing

Contributions are what make the open source community such an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated. For detailed contributing guidelines, please see CONTRIBUTING.md.

Copyright Notice and License

Distributed under Apache 2.0 License. See LICENSE for more information. Copyright AGNTCY Contributors.

Recommend Servers
TraeBuild with Free GPT-4.1 & Claude 3.7. Fully MCP-Ready.
Zhipu Web SearchZhipu Web Search MCP Server is a search engine specifically designed for large models. It integrates four search engines, allowing users to flexibly compare and switch between them. Building upon the web crawling and ranking capabilities of traditional search engines, it enhances intent recognition capabilities, returning results more suitable for large model processing (such as webpage titles, URLs, summaries, site names, site icons, etc.). This helps AI applications achieve "dynamic knowledge acquisition" and "precise scenario adaptation" capabilities.
EdgeOne Pages MCPAn MCP service designed for deploying HTML content to EdgeOne Pages and obtaining an accessible public URL.
Howtocook Mcp基于Anduin2017 / HowToCook (程序员在家做饭指南)的mcp server,帮你推荐菜谱、规划膳食,解决“今天吃什么“的世纪难题; Based on Anduin2017/HowToCook (Programmer's Guide to Cooking at Home), MCP Server helps you recommend recipes, plan meals, and solve the century old problem of "what to eat today"
CursorThe AI Code Editor
DeepChatYour AI Partner on Desktop
Serper MCP ServerA Serper MCP Server
MCP AdvisorMCP Advisor & Installation - Use the right MCP server for your needs
Tavily Mcp
BlenderBlenderMCP connects Blender to Claude AI through the Model Context Protocol (MCP), allowing Claude to directly interact with and control Blender. This integration enables prompt assisted 3D modeling, scene creation, and manipulation.
Visual Studio Code - Open Source ("Code - OSS")Visual Studio Code
Baidu Map百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Amap Maps高德地图官方 MCP Server
WindsurfThe new purpose-built IDE to harness magic
Playwright McpPlaywright MCP server
AiimagemultistyleA Model Context Protocol (MCP) server for image generation and manipulation using fal.ai's Stable Diffusion model.
ChatWiseThe second fastest AI chatbot™
TimeA Model Context Protocol server that provides time and timezone conversion capabilities. This server enables LLMs to get current time information and perform timezone conversions using IANA timezone names, with automatic system timezone detection.
MiniMax MCPOfficial MiniMax Model Context Protocol (MCP) server that enables interaction with powerful Text to Speech, image generation and video generation APIs.
Context7Context7 MCP Server -- Up-to-date code documentation for LLMs and AI code editors
Jina AI MCP ToolsA Model Context Protocol (MCP) server that integrates with Jina AI Search Foundation APIs.