Yomo

Install YoMo using the official install script, then verify the installation by checking the version.

curl -fsSL https://get.yomo.run | sh
yomo version

Create a new project directory and initialize it with a sample configuration. YoMo generates a my-agent.yaml configuration file and starter function code.

mkdir my-ai-agent && cd my-ai-agent
yomo init

Edit the generated my-agent.yaml to set your LLM provider, model, API key, and binding addresses. The zipper handles server coordination and http_api exposes the REST endpoint.

# my-agent.yaml
auth_token: "your-bearer-token"
zipper:
  host: 0.0.0.0
  port: 9000
http_api:
  host: 0.0.0.0
  port: 9001
llm_providers:
  - model: "gpt-4o"
    base_url: "https://api.openai.com/v1"
    api_key: "sk-xxxx"

Start the YoMo server with your configuration file. It will begin listening for incoming agent requests and function deployments.

yomo serve -c my-agent.yaml

In a separate terminal, deploy a function as a serverless tool that the LLM agent can call. Functions are deployed by name and connect to the running server.

yomo run -n get-weather

Configure your MCP client to connect to the YoMo server's HTTP API endpoint, making the deployed tool functions available to Claude or other AI models.

{
  "mcpServers": {
    "yomo": {
      "command": "npx",
      "args": ["yomo"],
      "env": {
        "YOMO_SERVER_URL": "http://localhost:9001"
      }
    }
  }
}