Think inside your AI world.
How do I connect Stripe to Claude?
Stripe ships an official MCP server, so this is one of the cleanest connections in the ecosystem: add it to Claude, authorise, and your live revenue surface is readable in chat.
Direct answer: Stripe publishes an official MCP server at mcp.stripe.com (around twenty-three tools). In Claude, open Settings → Connectors → add connector, point it at Stripe's endpoint and authorise; in Claude Code, a single 'claude mcp add' command does the same. From that moment Claude can read your account information, balances by currency, customers, invoices, subscriptions, payment intents, charges and disputes — with money-moving actions gated behind confirmation.
What the connection gives you
Ask about this month's charge volume, an invoice's status, a customer's subscription history or the current dispute count and the answer comes from your live account rather than the model's guesswork. Stripe keeps writes that move money behind explicit confirmation, which is the right default for a payments surface.
This is genuine, useful access, and if raw reach is all you need, Stripe's own connector completes the job.
The question the raw read leaves open
Suppose you ran a launch discount and ratified a rule first: refund rate stays under 2% across the promotion, or the pricing experiment stops — set because refunds, not top-line, sank the last one. Connected Stripe reports refunds at 2.6% happily and neutrally, because the 2% line and its history live in your head, not in your Stripe account.
The same connection, measured
Read through Unl, that Stripe figure arrives against the rule: over your 2% ceiling — the line you set after the January promotion — so by your own call, the experiment stops. Same live data, same one-way read; the difference is that your ratified criterion travels with it, so the answer is a verdict you authored rather than a rate you must remember to check.
Connecting Stripe to Claude takes minutes and returns your numbers; reading Stripe through Unl returns your numbers already held against the financial rules you set — met or breached, with your reasoning attached.
Reads through Unl arrive with measured context — in the presence of the decisions you’ve already settled. The reach lane is live: one box, paste anything. If it speaks MCP, Unl can reach it. Readings arrive unprompted, the data beside the criterion; Unl is a courier, not a warehouse, and keeps only your keys and the frame.
Read further
Questions people ask
What is the Stripe MCP server?
Stripe's official Model Context Protocol endpoint at mcp.stripe.com — roughly twenty-three tools spanning account info, balance by currency, customers, invoices, subscriptions, payment intents, charges, disputes, coupons and object search. Anything that moves money sits behind a confirmation gate.
Can I connect Stripe to ChatGPT the same way?
Yes — ChatGPT supports MCP connectors too, and Stripe's server speaks the same protocol to any MCP client. The phrasing people search — 'connect Stripe to ChatGPT' — resolves to the same endpoint and the same read surface.
Do I need Unl to connect Stripe to Claude?
No. Stripe's own server gives Claude full read access on its own, and this page's steps need nothing else. Unl adds the measured layer: the same read arriving inside the financial rules you ratified, so revenue questions come back as verdicts against your own thresholds.
What would I ratify in Unl for Stripe revenue?
The rules you already apply by hand, written down with the reason attached. Say your hire gate is MRR at or above £8k for two consecutive months, and the two-month part is the bit that carries the weight, because it stops one good spike reading as a trend. Ratified with that reasoning, the next Stripe read answers 'can I afford this?' against your rule — met, or met once and needing a second month — rather than handing you a figure to interpret.
Does Unl change or store my Stripe data?
Not stored: what Unl reads is quarantined for the turn and never kept — no query text and no read content reaches our telemetry. Not changed by a read, either — a read only reads. Where a server exposes a tool that writes, Unl can call it, because the socket exposes a server's full surface rather than a chosen subset; but every write carries your explicit in-turn gesture, verified at the choke point, and a tool that does not declare itself read-only is treated as a write and gated by default. Stripe keeps anything that moves money behind its own confirmation gate as well.
What this is
Think inside your AI world — you stay in command
Unlimitless (Unl to friends) holds what you've settled, reads what your tools are showing, and catches what's changed out in the world — and hands your AI whatever bears on the work, the moment it's needed, without you asking. The right thing, in front of the model, unprompted, with you in command of the call. So you keep moving toward what you set out to build, on top of everything you've already decided.
It plugs into Claude, Claude Code, ChatGPT and Cursor as an MCP connector. Quick to connect, in a couple of steps.
Unlimitless is open now to invited Alpha. Apply for the Beta waitlist to come in ahead of the full launch:
Alpha is invite-only · free at launch.