Imagine telling your coding agent to build and deploy a web app, and it handles everything from creating a Cloudflare account to registering a domain and getting an API token—all without you touching a dashboard. That's now reality. Through a new protocol co-designed with Stripe, agents can provision Cloudflare on behalf of users. They create accounts, start paid subscriptions, register domains, and return an API token for instant deployment. Humans only need to grant permission and accept terms of service. This eliminates manual steps like copying tokens or entering credit card details, letting agents go from zero to production in one shot. Below are the key questions about this groundbreaking capability.

What exactly can coding agents now do on Cloudflare?

Agents can perform every task a human customer would: create a Cloudflare account, activate a paid subscription, register a new domain, and obtain an API token to deploy code immediately. No prior Cloudflare setup is required—the agent starts from scratch. For example, using the Stripe Projects CLI, an agent can initialize a project, build an app, and deploy it to a freshly registered domain. The entire flow happens without the user visiting the Cloudflare dashboard, copying tokens, or manually entering billing details. Even payment methods can be added on the fly when the agent prompts the user. This means a single command like stripe projects init followed by a build prompt can result in a live, production application running on a new domain under a new Cloudflare account.

How does the new Cloudflare-Stripe protocol work?

The protocol was co-designed with Stripe and is part of the Stripe Projects launch. When a user installs the Stripe CLI with the Stripe Projects plugin and logs in, they can start a new project with stripe projects init. Their agent then takes over: it checks if the user's email has an existing Cloudflare account. If it does, a typical OAuth flow grants agent access. If not, Cloudflare automatically provisions a new account. The agent then builds the application, registers a domain via the Stripe Projects CLI, and sets up billing. Crucially, the agent pauses for human approval when needed—for example, if no payment method is linked. The result is a fully deployed app on a new domain, all managed by the agent. This protocol is designed so that any platform with signed-in users can integrate with Cloudflare similarly, offering zero friction for end users.

What happens if the user doesn't already have a Cloudflare account?

If the email used with Stripe has no associated Cloudflare account, Cloudflare automatically provisions one for the user and their agent. This is a seamless background process—no sign-up forms or emails to confirm. The agent receives the new account credentials and proceeds with setting up subscriptions and domain registration. The user only needs to accept Cloudflare's terms of service (a one-click action). This removes a major barrier because developers no longer need to create accounts manually or remember passwords. Combined with the Stripe integration, the entire onboarding happens silently, allowing the agent to focus on building and deploying. This automation is a game-changer for developers who want to iterate quickly without interrupting their flow to handle account creation.

How are payments and terms of service handled during this process?

Human oversight is maintained for two critical steps: accepting Cloudflare's terms of service and providing a payment method. When the agent needs to start a paid subscription or register a domain, it prompts the user to review and accept the terms. If the Stripe account lacks a linked payment method, the agent will ask the user to add one—for example, by prompting for a credit card. Once granted, the agent continues automatically. This ensures that while the agent does the heavy lifting, the human remains in control of legal and financial commitments. There is no need for manual dashboard navigation or token copying. The entire permission process is streamlined into clear prompts, making it both secure and efficient. After approval, the agent receives the API token and deploys the application to production without further human steps.

What are the key benefits for developers and startups?

The main benefit is radical simplification: developers can go from idea to production in one command. No more juggling tabs to create accounts, generate API tokens, or set up billing. This is especially powerful for startups and prototyping. Cloudflare is sweetening the deal by offering $100,000 in Cloudflare credits to all new startups that incorporate using Stripe Atlas. This means new ventures can get significant free infrastructure while enjoying zero-friction agent deployment. Additionally, the agent can leverage Cloudflare's Code Mode MCP server and Agent Skills to build and deploy even more effectively. By eliminating manual setup, developers can focus on product development, and agents can autonomously handle the entire lifecycle—from account creation to production hosting—saving hours of repetitive work.

Can other platforms integrate with Cloudflare the same way Stripe does?

Yes, the new protocol is designed to be generic. Any platform with signed-in users can adopt this integration to allow agents to provision Cloudflare on behalf of those users. The protocol handles account creation, subscription management, domain registration, and API token generation in a backend fashion. Platforms simply need to implement the same endpoints and callbacks that Stripe uses. This opens the door for other developer tools, CI/CD services, and low-code platforms to offer instant Cloudflare deployment. For end users, the experience remains frictionless—they only see permission prompts and terms acceptance. This extensibility is a significant step toward a future where agents can interact with any cloud provider on behalf of users, with the same ease as a human using a dashboard.