Cloudflare Containers
Connect AI tools to Cloudflare Containers to spin up sandbox development environments. The Cloudflare Containers MCP server gives AI assistants programmatic access to container-based sandboxes through the Model Context Protocol (MCP).
1. Overview
Cloudflare Containers is a remote, vendor-hosted MCP server provided by Cloudflare. You connect to it from Cequence AI Gateway; the server uses Streamable HTTP (the standard transport for remote MCP per Cloudflare’s MCP docs).
- Server URL:
https://containers.mcp.cloudflare.com/mcp - Transport: HTTP (Streamable HTTP)
- Hosted by: Cloudflare
2. Supported authentication types
| Type | Supported | Notes |
|---|---|---|
| OAuth 2.0 | Yes | Required. Uses Dynamic Client Registration (DCR); sign in with your Cloudflare account. |
| API key | No | Not used for this remote MCP server. |
When you add Cloudflare Containers in Cequence AI Gateway, authentication is handled via OAuth 2.0 with Dynamic Client Registration. You sign in with your Cloudflare account and grant access during the gateway flow.
3. What can you do with this MCP server
With the Cloudflare Containers MCP server, you can:
- Spin up sandboxes — Create and manage sandbox development environments on Cloudflare’s network.
- Run and manage containers — Use natural language to start, stop, or inspect containerized environments.
- Run and test code — Execute code in a secure, isolated environment when your MCP client does not have a built-in dev environment (e.g. claude.ai).
- Validate in real time — When building or generating application code, let the AI run commands and validate assumptions in real time inside the sandbox.
4. Prerequisites
Before adding Cloudflare Containers in Cequence AI Gateway, ensure you have:
- Access to Cequence AI Gateway (e.g. beta.aigateway.cequence.ai)
- A Cloudflare account with access to Containers
- A modern browser to complete the OAuth authorization flow
- For OAuth authentication: an auth app with client credentials (client ID and client secret) in your Cloudflare (vendor) account, unless the server supports Dynamic Client Registration (DCR).
5. Example workflows
- “Create a new sandbox environment with Node 20.”
- “List my running container sandboxes and their status.”
- “Stop the sandbox with ID [id].”
6. Connecting MCP server from Cequence AI Gateway
- Log in to Cequence AI Gateway.
- Choose your tenant.
- Go to App catalogue.
- Filter by Remote MCP server.
- Search for Cloudflare Containers and then select it.
- Click Create MCP server.
- Choose auth method. If OAuth, you need an auth app with client credentials in your vendor account (see Prerequisites).
- Complete the setup as prompted, select tools, and deploy.
Use the generated MCP server URL in your client as described in the Client Configuration docs. For detailed UI steps and screenshots, see Create a third-party MCP Server.
7. Additional information
- Transport: Streamable HTTP. See Transport.
- Timeout: 30-second timeout for requests.
- Cloudflare MCP: Model Context Protocol (MCP), Build a Remote MCP server.
- Official repository: mcp-server-cloudflare.