For AI agents

Let your assistant ask, politely.

Every Yorn link is also an endpoint for AI agents. They send a structured offer and get a single TRUE or FALSE back, with the same privacy and fair-play rules as people.

Read the docs Get a link to test

Protocol

Three calls. That's the whole thing.

list_gates()

Discover what a person is open to: categories and fields, never their limits.

{ "gates": ["work", "rent", "brands"] }

submit_offer()

Send a structured offer. Get TRUE or FALSE.

{ "decision": TRUE }

claim_channel()

After a TRUE, open the verified conversation.

{ "channel": "verified" }

Same rules as humans

One free attempt per gate every 30 days
Step-by-step guessing is detected and blocked
No reasons and no partial scores, ever

Three doors, one gate

People visit. Agents query. Systems call.

Assistants discover gates and ask on someone's behalf. They get the same single answer.

mcp://yorn.si/you

→ tools/call   mcp://yorn.si/lucia
{
  "name": "submit_offer",
  "arguments": {
    "gate": "rent",
    "rent_eur": 1450,
    "occupants": 2,
    "lease_months": 12
  }
}
← { "decision": TRUE, "claim_token": "clm_9xK2f…" }