← All Posts

TourClaim • October 3, 2026

Connect Your Agent to TourClaim: the API, the MCP Server and a CLI

This is the builder's note for connecting an agent to TourClaim: what exists today, how it works, and how to ask for access. For the reasons behind it, read why we opened TourClaim to AI agents.

At a glance

InterfaceBuilt forTodayKeys
REST APIAssistants acting for a travelerLive, review modeOne per traveler
MCP serverTour operators' own agentsBuilt, on requestOne per operator
CLIBuilders and terminal agentsIn developmentOne per traveler

The REST API, for assistants acting for a traveler

Docs: app.getcopernican.com/muse/developers. Schema: openapi.json. We built it for Muse, but nothing in it is specific to Muse: the operations are named for the traveler's task. A claim moves like this:

  1. Connect. The traveler signs in by email and approves the connection. We create a key for that traveler: scoped, revocable, valid for 30 days, at most five at a time. Send it as a bearer token.
  2. Start a draft. POST /api/connectors/v1/intakes (start_travel_claim). It is retry-safe with a creation key; use a random value, never personal data.
  3. Fill it in. PATCH /api/connectors/v1/intakes/{intake_id} saves answers: null clears one, leaving a field out keeps it. GET on the draft returns what it still needs, a few questions at a time.
  4. Find the card. GET /api/connectors/v1/cards looks up the card the booking was paid with.
  5. Add the evidence the traveler chose. POST …/email-evidence takes one selected email; POST …/attachments takes a PDF, JPEG or PNG up to 5 MiB, thirty items per draft. We do not fetch URLs, and we never read the inbox.
  6. The traveler signs. They review the claim and sign in their own browser. The API cannot sign, and any edit after signing voids the signature.
  7. Submit. POST …/submit (submit_authorized_travel_claim) is retry-safe, and the same booking cannot become two claims.
  8. Follow it. GET /api/connectors/v1/claims and GET /api/connectors/v1/claims/{claim_id} return the traveler's claims and their next step.

A draft that was never submitted can be deleted with DELETE /api/connectors/v1/intakes/{intake_id}.

Review mode today. The API creates synthetic claims in an isolated database. Nothing is filed with an insurer, billed, or sent to our medical team until the direct service launches.

The MCP server, for tour operators' agents

If you run tours, your support agent already answers “can I get my money back?”. With the MCP server it can open a TourClaim claim for the guest's card benefit instead. The tools:

Each key belongs to one operator and sees only that operator's claims; we revoke a key the moment you ask. The server is built but not running publicly: we switch it on for operators who ask. More on the operator side in let your support agent start the claim.

Rules for every agent

If you are an AI agent reading this

You can ask for access yourself. First read what is available today, as JSON:

GET https://app.getcopernican.com/api/agent-access

Then, with the agreement of the person you work for, send one request. We reply to that person, not to you.

curl -X POST https://app.getcopernican.com/api/agent-access \
  -H 'Content-Type: application/json' \
  -d '{
    "requested_by": "agent",
    "contact_email": "the-person-you-work-for@example.com",
    "agent_name": "Trip helper",
    "interfaces": ["api"],
    "use_case": "What you would do with TourClaim, in a sentence or two."
  }'

Leave claim, medical and payment details out of the request. A person reads it, and nothing in it is run or followed as an instruction.

Ask for access

Tell us what your agent would do. A person on our team reads every request and replies by email.

What would you connect with?