On June 17, 2026, Google announced Agentic Resource Discovery (ARD) — an open specification for how AI agents find tools, APIs, and other capabilities across the web. Storylayer is among the first publishing platforms to ship a live ARD catalog.
The problem ARD solves
Today, connecting an AI agent to a new service means manual configuration: find the URL, read the docs, paste API keys, hope the client supports the protocol. That breaks down as the number of agentic services explodes.
ARD standardizes discovery before invocation. An agent asks "what can help me schedule Instagram posts?" and a registry returns matching capabilities with trust metadata — then the agent connects via MCP, REST, or whatever native protocol the service uses.
What Storylayer publishes
Every Storylayer deployment exposes these well-known endpoints on app.storylayer.ai:
| Endpoint | Purpose |
|---|---|
| /.well-known/ai-catalog.json | ARD manifest with semantic search queries |
| /.well-known/mcp.json | MCP Server Card for pre-connection discovery |
| /.well-known/openapi.json | Full REST API spec |
| /.well-known/server.json | MCP Registry entry |
| /.well-known/agent-quickstart.md | Plain-text onboarding for agents |
Each catalog entry includes representativeQueries — natural language intents like "connect Claude to Instagram and schedule posts from Airtable" — that power semantic search in federated registries.
Why this matters for Storylayer users
If you use Storylayer via Claude, ChatGPT, or Cursor today, you already benefit from our MCP server and OAuth 2.1 setup. ARD is the next layer: it means future AI products can discover and connect to Storylayer without you pasting URLs.
As Google Agent Registry, Hugging Face Discover, and enterprise agent platforms adopt ARD, Storylayer will appear in their catalogs automatically.
What's next
- Submit our catalog as a reference implementation to the ARD spec repository
- Register with the MCP Registry
- Expand OpenAPI coverage as the API grows
Read the full implementation guide in our developer docs or fetch the live catalog:
curl https://app.storylayer.ai/.well-known/ai-catalog.json