See it work — 60 seconds, nothing to install

When someone asks an AI to book a trip, does your business show up — with today's real price?

Most AI assistants can't see your actual availability — they guess, or send the customer to whoever integrated first. Archstone connects your real prices and inventory directly to ChatGPT, Claude and Gemini, so they can look it up and get it right, instead of making something up.

Try it right now — ask a real question Open-source core on GitHub

You describe what you offer, once. No integration team, no per-model rebuild.

You own the data

Answers come only from your own portfolio — every supplier you signed. No scraping, no external sources, no surprises.

We stay neutral

We pass your information to the AI exactly as you gave it. We never rank your offers against anyone else's, and we never decide what the customer sees first — the assistant does, for the traveler.

The principle

Your business capabilities are the source of truth. Everything else is generated.

— you describe your business once, and Archstone builds the rest

Why it hasn't happened yet

Every AI is its own integration project. Your APIs were built for people, not agents. So the channel that matters most is the one you can't staff for.

A CTO knows the AI channel is coming, but can't spare three developers for months of work per model. Archstone turns that whole effort into a single description — written once, kept in sync for you.

Declare once, reach all

One capability contract exposes you to every model at once. The next assistant that matters is a config line, not a project.

Live, not illustrated

Try it in under a minute — any device, nothing to install.

tourism_search is a real, working example — already built, hosted live. This works the same way on Mac, Windows, the web, or your phone: no config file, no terminal, no sandbox account.

Open Claude — claude.ai, the Claude Desktop app, or the mobile app.
Go to Settings (or Customize) → Connectors.
Click Add custom connector, paste the URL below, click Add:

Connector URL — paste this

https://demo.archstone.dev/mcp

Then ask about a trip — a destination, a date range, a budget. Claude looks it up against a real backend and answers from what it finds: the exact scenario below, live. Works on Free, Pro, Max, Team and Enterprise plans (Free accounts get one custom connector — this is it).

Prefer a terminal? Claude Code — one line, same on Mac/Windows/Linux

claude mcp add --transport http archstone-tourism https://demo.archstone.dev/mcp

For transparency: this demo's backend returns fixed, deterministic data — not a live operator's real inventory — so you can see the exact shape of what comes back. Same query, same result, every time. Curl it yourself:

Raw response — try it

curl -s -X POST https://demo.archstone.dev/v1/search \
  -H "content-type: application/json" \
  -d '{"destination":"Rome"}'

What it returns

{
  "stays": [
    { "id": "stay-5-0", "name": "Riverside Lodge — Rome", "location": "Rome",
      "pricePerNight": 225, "rating": 4.1 },
    { "id": "stay-2-1", "name": "The Olive Court — Rome", "location": "Rome",
      "pricePerNight": 278, "rating": 4.5 },
    { "id": "stay-9-2", "name": "Harbor House — Rome", "location": "Rome",
      "pricePerNight": 331, "rating": 4.9 }
  ]
}

Prefer everything on your own machine — model included? archstone serve speaks the same protocol over stdio, so a local model via Ollama or LM Studio calls it exactly like Claude does above — no cloud, no code changes. Copy-paste config in the open-source docs.

How it works

One description. Every model. Written once.

You describe a capability — what your business can do, in plain business terms, not code. Archstone turns that into whatever format each AI needs, and keeps it up to date. Three steps, and the last one repeats for free.

STEP 01

Describe

State the capability — its inputs, its result, who may call it. Business meaning, not endpoint plumbing.

STEP 02

Archstone builds it

It generates the connectors, policy and context layer. Nobody hand-writes integration code.

STEP 03

Every AI invokes

The same capability is discoverable and callable across models — new assistants added without touching your systems.

tourism.search — one declaration

# a capability is what you do, not how your API is wired
capability: tourism.search
description: Find places to stay matching a traveler's intent.
input:
  destination: location
  dates:       date-range    # flexible — prices vary by day
  travelers:   party
  budget:      money         # optional
output:
  stays: collection<Stay>   # from your portfolio only

Already comfortable building an MCP server yourself? Then the work was never the first server — it's the fifth one, after ChatGPT's, Gemini's and whatever comes next each want it shaped slightly differently, and your API changes underneath all of them. Archstone keeps every one of those in sync from a single description, so that maintenance disappears instead of multiplying.

Neutral hosting

Don't want to run your own node? We'll host the bridge.

Some teams would rather not manage their own server — that's fine. Hosting is optional, and it's pure utility: no bidding wars that rank one business above another. We host the technical bridge; you keep absolute control over where it leads.

Available on request while we build self-serve hosting — write to us and we'll set it up for your capability.

In practice — live, not illustrated

A real customer asked. A real assistant answered — from real inventory.

This isn't a mockup. It's Meșterul, the AI assistant running today on artvinci.ro, a real custom-framing business. It answers only from ArtVinci's actual catalog and pricing — computed live by their own backend, through the exact kind of capability shown in "How it works."

Captured live, 2026-07-25 — real question, real backend, real prices. Asked in Romanian on the live site; shown here in English.

Customer, on artvinci.ro

How much does a frame cost for a 40×50cm painting?

→ invokes framing.estimate-frame-price · ArtVinci's real catalog

Meșterul — ArtVinci's assistant

For a painting sized 40×50cm, here are estimates from a few of our profiles: AV22, antiqued gold-blue — 63.80 RON; AV8, brown wood with patina — 79.39 RON; AV12, gold wood — 563.08 RON. These prices are calculated for the exact dimensions given — tell me a style, material or color you prefer and I can narrow it down further.