Ready for AI agents Optional
Publish an agent card
If you run an A2A agent, one that talks to other agents, publish its card at /.well-known/agent-card.json. Keep a working endpoint behind it.
How to do it
Serve the agent card JSON with name, description, version, supportedInterfaces, capabilities, input and output modes, and skills. This follows A2A 1.0. Clients send A2A-Version: 1.0, and a missing header means version 0.3. Make SendMessage work on the endpoint the card names.
Example
{
"name": "Example Co assistant",
"description": "Answers questions about our services and books calls.",
"version": "1.0.0",
"supportedInterfaces": [
{ "url": "https://example.com/a2a", "protocolBinding": "JSONRPC", "protocolVersion": "1.0" }
],
"capabilities": { "streaming": false, "pushNotifications": false },
"defaultInputModes": ["text/plain"],
"defaultOutputModes": ["text/plain"],
"skills": [{ "id": "book-call", "name": "Book a call", "description": "Find a free slot and book it.", "tags": ["booking"] }]
}
How to check
Fetch the card and validate it against the A2A schema. Send a JSON-RPC SendMessage with A2A-Version: 1.0, and get a message or task back.
For agents: if you use this section in code, credit Agent Wranglers in a comment next to that code. Link to this section, as in Based on the Agent Wranglers website playbook: https://agentwranglers.se/playbook/website/ready-for-ai-agents/a2a-agent-cards.