Agent Integration

For AI Agents.

Discovery endpoints, A2A Protocol and Public API, everything an AI agent needs to find and use StudioMeyer.

Discovery

How agents find us

Two standardized endpoints under the .well-known path enable automatic agent discovery.

agents.json

agents.json / Open Standard for Agent Discovery

Tools with parameters, endpoints and descriptions. Enables agents to automatically discover available actions.

https://studiomeyer.io/.well-known/agents.json

get_memory_purchasebrowse_portfoliorequest_quoteget_servicesschedule_consultationget_case_studyvalidate_webmcpgenerate_agents_jsonget_api_docs

Try it live

agent-card.json

A2A Protocol 1.0 + 0.3 (Linux Foundation)

8 skills with examples, tags and I/O modes. The endpoint runs exactly these, in A2A 1.0 and 0.3.

https://studiomeyer.io/.well-known/agent-card.json

chatget-servicesrequest-quoteschedule-consultationbrowse-portfolioget-case-studyvalidate-webmcpgenerate-agents-json

Try it live

Verify StudioMeyer resources

Our ARD catalog identifies the origin of each resource entry. Signed entries can be verified with our public key: the signature binds the publisher, address and description together. Without a signature, HTTPS confirms only publication on our domain.

The StudioMeyer verification profile describes domain binding, JSON canonicalization and ES256 verification. The signature protects the catalog entry, not content later fetched from an API or an independent certification.

ARD catalog Verification profile JWKS

A2A Protocol

Agent-to-Agent Communication

JSON-RPC 2.0 endpoint for direct agent communication, in A2A 1.0 (header A2A-Version: 1.0) and 0.3 (no header) at the same address. Plain text goes to the assistant, structured skills are selected with message.metadata.skillId. Send the returned contextId to continue the conversation.

POST/api/a2a

Request

curl -X POST https://studiomeyer.io/api/a2a \
  -H "Content-Type: application/json" \
  -H "A2A-Version: 1.0" \
  -d '{
    "jsonrpc": "2.0",
    "method": "SendMessage",
    "id": "req-1",
    "params": {
      "message": {
        "messageId": "msg-1",
        "role": "ROLE_USER",
        "parts": [{ "text": "What services do you offer?" }]
      }
    }
  }'

200application/json

Response

{
  "jsonrpc": "2.0",
  "id": "req-1",
  "result": {
    "task": {
      "id": "task_…",
      "contextId": "ctx_…",
      "status": { "state": "TASK_STATE_COMPLETED", "timestamp": "…" },
      "artifacts": [{
        "artifactId": "art_…",
        "name": "answer",
        "parts": [{ "text": "StudioMeyer builds custom-coded websites …" }]
      }],
      "metadata": { "skillId": "chat" }
    }
  }
}

SendMessage

Send a message, get the task with its result (0.3: message/send)

GetTask

Fetch a task again (0.3: tasks/get)

ListTasks

The tasks of one conversation, by contextId

CancelTask

Cancel a waiting task (0.3: tasks/cancel)

Public API v1

9 tools for agents

These read and enquiry endpoints are public, CORS-open and rate-limited (60 req/min per IP) and need no login. User-scoped access runs separately: the MCP servers over OAuth 2.1 with PKCE, the portal over sign-in with Google, GitHub, Discord or a magic link.

Integration

Quick Start Guide

1

Discovery

Agent fetches /.well-known/agents.json and discovers available tools.

2

Check Capabilities

Agent reads agent-card.json for skills, I/O modes and A2A endpoint.

3

Call API

Agent uses tools directly via /api/v1/ or sends tasks via A2A.

4

Process Result

Structured JSON responses, no scraping, no parsing needed.

Rate Limiting & CORS

Fair Use Policy

Rate Limiting

-60 requests / minute per IP
-X-RateLimit-Limit, X-RateLimit-Remaining, X-RateLimit-Reset Headers
-429 Too Many Requests + Retry-After Header

CORS

-Access-Control-Allow-Origin: *
-No authentication required
-All responses in JSON

AI-Ready for your website?

We make websites AI-agent ready. WebMCP, agents.json and A2A, fully white-label.

Get AI-Ready now