Skip to main content

Host your Agent

Overview

You can self-host your Agent using any hosting stack.

The following guide covers the essential steps you need to take after creating a Warden Agent, but the exact flow may vary depending on your hosting stack.

1. Edit the Agent code

First, configure how your Agent determines its external base URL.

In src/agent.ts, edit the following code:

const AGENT_URL = process.env.AGENT_URL || BASE_URL;

Set process.env.AGENT_URL to the publicly accessible URL of your deployed service. The exact environment variable depends on your hosting provider.

tip

Many platforms expose this variable automatically, for example: RENDER_EXTERNAL_URL (Render), VERCEL_URL (Vercel), and so on.

2. Publish the code

Now you need to publish your Agent on GitHub:

  1. Keep your environment variables from the .env file in a safe space. This includes the Agent API key for authorization, the LLM API key, and other parameters.

  2. Delete the .env file.

    warning

    Never expose your API keys on GitHub.

  3. Create a new repository on GitHub.

  4. Push your Agent code to the repository.

3. Deploy your Agent

To set up a hosting service, take these steps:

  1. Create a new web service with your preferred hosting provider.

  2. In the service configuration, set the required environment variables:

    • AGENT_API_KEY: The Agent API key for authorization
    • OPENAI_API_KEY/other: The API key for OpenAI or other LLM
    • OPENAI_MODEL/other: The preferred LLM model

    If your Agent uses x402 payments, you'll need to add more variables. See Configure x402 payments.

  3. Set a build command:

    npm build
  4. Set a start command:

    npm run start
  5. Connect your repository with the hosting service.

  6. Trigger the deployment if it doesn't start automatically.

Congratulations! Now your Agent and the supported API endpoints will be available on a public URL provided by your hosting service.

4. Verify deployment

Chat using the UI

The fastest way to make sure your Agent is accessible is opening the public URL provided by the hosting.

If everything is fine, you'll be able to chat with your Agent through the user interface provided by Warden Code:

The user interface for chatting with Agents, provided by Warden Code

Chat using the CLI

To chat with your Agent through the CLI, take these steps:

  1. Navigate to your project's root directory and initiate Warden Code:

    warden
  2. Enter the chat mode and specify the Agent's public URL provided by the hosting:

    /chat PUBLIC_URL
  3. Chat with your Agent. Warden Code automatically detects whether the agent supports A2A, LangGraph, or both, and prompts you to choose when multiple protocols are available.

  4. To exit the chat mode, type this:

    /exit

Chat using the API

Every new Agent is immediately accessible through A2A and LangGraph server endpoints exposed by Warden Code.

tip

For a full list, see A2A endpoints and LangGraph endpoints.

After deploying your Agent, you can try any of these endpoints. For example, you can prompt the Agent using the A2A POST endpoint with the send method.

In the request below, specify the following:

  • PUBLIC_URL: Your public URL
  • AGENT_API_KEY: Your Agent API key for authentication:

POST PUBLIC_URL
Headers: Content-Type: application/json
Authorization: Type: Bearer Token, Token: AGENT_API_KEY
Body:

{
"jsonrpc": "2.0",
"id": "",
"method": "message/send",
"params": {
"message": {
"role": "user",
"parts": [
{
"kind": "text",
"text": "What can you do?"
}
],
"messageId": ""
},
"thread": {
"threadId": ""
}
}
}

If everything is fine, you'll receive a response including your prompt, assistant's reply, and other data:

{
"jsonrpc": "2.0",
"result": {
"id": "task-2",
"context_id": "da79d131-143c-4154-b617-c25945774648",
"status": {
"state": "completed",
"timestamp": "2026-02-09T08:03:07.107Z"
},
"kind": "task",
"history": [
{
"role": "user",
"parts": [
{
"kind": "text",
"text": "What can you do?"
}
],
"messageId": "",
"kind": "message",
"message_id": "5646cc50-ae72-40a9-87e6-99477e050a4f"
},
{
"role": "agent",
"parts": [
{
"kind": "text",
"text": "I can assist with a variety of tasks, including but not limited to:\n\n1. Answering questions and providing information on a wide range of topics.\n2. Assisting with problem-solving and brainstorming ideas.\n3. Offering writing support, including proofreading, editing, and generating text.\n4. Providing summaries and explanations of complex concepts.\n5. Helping with language translation and learning.\n6. Offering recommendations for resources, books, or tools.\n7. Engaging in casual conversation and providing companionship.\n\nIf you have a specific request or need assistance with something else, feel free to ask!"
}
],
"kind": "message",
"message_id": "d187905d-2ded-41e5-87ae-db02ee011d88"
}
]
},
"id": ""
}

Check the Agent Card

As part of the A2A protocol support, Warden Code exposes an endpoint for the A2A Agent Card, which advertises your Agent's skills, allowing other Agents and clients to discover it.

To check the Agent Card accessibility, replace PUBLIC_URL with the public URL and run this:

PUBLIC_URL/.well-known/agent-card.json

The card will display your Agent's name and capabilities, along with other information:

{
"name": "general-test",
"description": "A helpful AI agent named general-test",
"url": "PUBLIC_URL",
"version": "0.1.0",
"capabilities": {
"streaming": true
},
"defaultInputModes": ["text/plain"],
"defaultOutputModes": ["text/plain"],
"skills": []
}
tip

Most likely, the url field will display the local host URL instead of the public URL. To fix this issue, update the Agent Card.

Next steps

Now you can do the following: