Cloudflare

Cloudflare

Deploy, configure & interrogate your resources on the Cloudflare developer platform (e.g. Workers/KV/R2/D1)

3,157
Stars
299
Forks
20
Releases

Overview

Model Context Protocol (MCP) is a standardized protocol for managing context between LLMs and external systems. This repository hosts multiple Cloudflare MCP servers that let an MCP client connect to Cloudflare services using natural language to configure, query, and modify resources across your Cloudflare account. The servers enable clients to read account configurations, process information, generate data-driven suggestions, and apply changes across a wide range of Cloudflare services—from application development to security and performance. Interactions can occur over the streamable-http transport at /mcp, or via the SSE transport (deprecated) at /sse. The repository includes dedicated servers for various Cloudflare tools (Documentation, Workers Bindings, Workers Builds, Observability, Radar, Sandbox Container, Browser Rendering, Logpush, AI Gateway, AutoRAG, Audit Logs, DNS Analytics, DEX, Cloudflare One CASB, GraphQL), each with its own domain and endpoints. It also provides guidance on accessing remote MCP servers, using them with OpenAI’s Responses API, configuration via mcp-remote, troubleshooting, paid feature notes, and contribution guidelines.

Details

Owner
cloudflare
Language
TypeScript
License
Apache License 2.0
Updated
2025-12-07

Features

Remote MCP server access from MCP clients

Provides a centralized set of Cloudflare MCP servers that MCP clients can connect to over HTTP, enabling cloud services automation via natural language.

Support for streamble-http transport

The MCP servers support the streamble-http transport via /mcp, enabling streaming interactions between LLMs and Cloudflare services.

SSE transport (deprecated)

Supports SSE transport via /sse, though it is deprecated.

Dedicated Cloudflare MCP servers

Includes multiple servers each focused on a Cloudflare service (Documentation, Workers Bindings, Workers Builds, Observability, Radar, Sandbox Container, Browser Rendering, Logpush, AI Gateway, AutoRAG, Audit Logs, DNS Analytics, DEX, Cloudflare One CASB, GraphQL).

OpenAI Responses API integration guidance

Provides instructions for using Cloudflare MCP servers via OpenAI's Responses API, including API token scopes.

mcp-remote configuration sample

Includes an example configuration for mcp-remote to specify which servers your client can access.

Troubleshooting context-length issues

Offers guidance to avoid Claude interruptions due to context-length limits by simplifying queries and breaking requests into smaller tool calls.

Contributing and paid features notes

Information on contributing and notes about paid Cloudflare Workers plans for certain features.

Audience

DevelopersConnect and automate Cloudflare services via MCP using natural language interactions.
Cloudflare administratorsManage configurations, monitor analytics, and apply changes across multiple Cloudflare services.
MCP clientsAccess Cloudflare MCP servers remotely, query configurations, and issue actions via natural language.

Tags

cloudflaremcpserverremoteworkersobservabilitydns-analyticsgraphqlbrowser-renderinglogpushai-gatewayautoragauditlogsdex-analysiscasbsandboxdocumentationopenaimcp-remote