How to Connect Stripe to an AI Agent: Step-by-Step Guide
· DoubleX
Quick answer
To connect Stripe to an AI agent, open the Connectors screen in DoubleX Agentes, choose Stripe, and authorize access. Then pick which actions the agent can run (read-only, or also creating links and invoices), choose who can use the connection, and test it with a simple first request.
Connecting Stripe to an AI agent takes a few minutes, but the choices you make along the way matter: which account you authorize, which actions the agent may run, and who can use the connection. This guide covers the setup only. For ideas on what to ask once it works, see the guide on Stripe with an AI agent: example requests and routines.
What can an AI agent do with Stripe?#
With the Stripe connector, an agent can look up customers, payments, and subscriptions, and it can create payment links and invoices. An AI agent is an assistant that uses your business systems to do work from a plain-language request, instead of only answering questions.
That means you can ask in your own words, for example "Which customers have an active subscription?", and the agent uses Stripe's tools to find the answer. Creating links and invoices is optional, and you decide whether to allow it.
What do you need before connecting?#
You need a DoubleX Agentes account, access to the Stripe account you want to connect, and a decision about how much the agent should be allowed to do.
- A DoubleX Agentes account. Creating one and running your first diagnostic is free, with no credit card.
- Login access to the Stripe account, with enough permission to authorize a connection.
- An agent to use it. You can create one from the diagnostic, which turns a short conversation about your routine into a team of agents.
- A rule of thumb: start with read-only. You can allow more later.
How do you connect Stripe on the Connectors screen?#
You connect Stripe from the Connectors screen by selecting it from the list and authorizing access with your Stripe login. Here is the flow in DoubleX Agentes:
- Sign in and open the Connectors screen.
- Find Stripe in the finance category and select it.
- Authorize access with your Stripe login when prompted.
- Confirm that Stripe now shows as connected.
- Open the agent that should use it on the Agents screen and enable the Stripe connector for that agent.
Credentials and tokens for connections are stored encrypted (AES-256-GCM), and each company has its own separate database. For details on how connectors work, see the Connectors documentation.
How do you control what the agent can do: read-only or also links and invoices?#
For each agent, you choose which systems it uses and which actions it can run in each one. For Stripe, that means you can leave the agent read-only, or also allow it to create payment links and invoices.
| Setup | What the agent can do | Good for |
|---|---|---|
| Read-only | Look up customers, payments, and subscriptions | Questions, checks, and reports |
| Read plus create | Also create payment links and invoices | Billing tasks you want to hand off |
Actions that change data can stay turned off. Even when they are on, the agent asks for confirmation before sending, deleting, or changing anything in your systems, and it shows the step-by-step of what it did.
Who can use the connection: personal, team, or company?#
A connection can be personal, shared with a team, or shared with the whole company. A personal connection only works for the person who made it, like your own email. A team or company connection lets other people's agents use the same Stripe access.
For finance data, a team connection is often the better fit, limited to the people who handle billing. Roles (owner, admin, member), teams, and sharing by person, team, or company help you keep this under control. Also remember that a scheduled routine runs with the permissions of whoever created it, so create routines with an account that has only the access needed.
How do you test with a first request and check the history?#
Test with a simple, read-only request, then open the task to see exactly what the agent did. Every request to an agent becomes a task on the Tasks screen, with the request, each tool the agent used, and the result, so you can see who asked, what was done, and when.
Copy one of these to start:
- "List the last 10 payments in Stripe and show the customer, amount, and status."
- "Which customers have an active subscription?"
- "Find the customer with the email [email protected] and summarize their payments."
- "Create a draft payment link for the Basic plan and ask me to confirm before sending."
The last example only works if you allowed creating links. If the answer looks wrong, open the task, check which Stripe tools were called, and adjust the agent's instructions or permissions. The Quickstart walks through the general flow, and the Introduction explains the main concepts.
Frequently asked questions#
Do I need to code to connect Stripe?#
No. You authorize Stripe on the Connectors screen and describe what you need in your own words. Developers who want more control can use the API, CLI, or MCP server.
Can I keep the agent from creating invoices?#
Yes. You choose which actions the agent can run in Stripe. Leave it read-only, and it can look things up but cannot create links or invoices.
Will the agent change anything without asking?#
No. Before sending, deleting, or changing something in your systems, the agent asks for confirmation and shows what it did step by step.
Is my Stripe data used to train AI models?#
No. Your company's content is not used to train models, and requests only go to AI providers that do not use what they receive for training.
Can I see what the agent did in Stripe?#
Yes. Each task keeps a full history: the request, every tool the agent used, and the result.
Related guides
- How to Connect Box to an AI Agent: Step-by-Step GuideLearn how to connect Box to an AI agent: set up the connector, limit it to read-only, choose who can use it, and check the history of your first request.
- How to Connect Jam to an AI Agent: Step-by-Step GuideLearn how to connect Jam bug reports to an AI agent in DoubleX Agentes: set up the connector, pick who can use it, limit permissions, and run a first test.