Remote MCP Server on Cloudflare
@tomtev
About Remote MCP Server on Cloudflare
No overview available yet
Config
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"remote-mcp-server-tomtev": {
"command": "npx",
"args": [
"nx",
"dev",
"remote-mcp-server"
]
}
}
}Tools
No tools detected
Fetch the live tool list by running this server in a temporary sandbox using the button above.
Overview
What is Remote MCP Server on Cloudflare?
A template for hosting a remote MCP (Model Context Protocol) server on Cloudflare Workers, complete with OAuth login. It is intended for developers who want to run MCP tools on a serverless platform and connect them to AI clients like Claude Desktop.
How to use Remote MCP Server on Cloudflare?
Clone the repository, install dependencies, and run npx nx dev remote-mcp-server for local development. Deploy by creating a KV namespace with npx wrangler kv namespace create OAUTH_KV, adding the namespace ID to wrangler.jsonc, and running npm run deploy. Connect clients (MCP Inspector or Claude Desktop) via SSE at the server’s /sse endpoint.
Key features of Remote MCP Server on Cloudflare
- Remote MCP server running on Cloudflare Workers
- OAuth login flow built in
- SSE (Server‑Sent Events) transport
- Local development with MCP Inspector support
- Direct integration with Claude Desktop via
mcp-remoteproxy
Use cases of Remote MCP Server on Cloudflare
- Deploying MCP tools on a serverless edge network
- Remotely calling MCP tools from AI assistants like Claude
- Prototyping and testing MCP servers locally before deploying
FAQ from Remote MCP Server on Cloudflare
How do I run the server locally?
Clone the repo, run npx nx dev remote-mcp-server in the ai directory, then open http://localhost:8787/ in a browser.
How do I connect the MCP Inspector to my local server?
Start the inspector with npx @modelcontextprotocol/inspector, set Transport Type to SSE and URL to http://localhost:8787/sse, then click “Connect”. You will be prompted to log in (any email/password works in the mock flow).
How do I deploy the server to Cloudflare?
Create a KV namespace with npx wrangler kv namespace create OAUTH_KV, paste the returned ID into your wrangler.jsonc file, then run npm run deploy.
How do I connect Claude Desktop to a remote instance?
Update your Claude Desktop config file to use the mcp-remote command and point it to your workers.dev URL (e.g., https://worker-name.account-name.workers.dev/sse). Restart Claude; a browser window will open for login.
How can I debug connection issues?
Try connecting directly from the command line with npx mcp-remote http://localhost:8787/sse. In rare cases, clearing the ~/.mcp-auth folder can resolve authentication problems. Restarting Claude may also help.
Frequently asked questions
How do I run the server locally?
Clone the repo, run `npx nx dev remote-mcp-server` in the `ai` directory, then open `http://localhost:8787/` in a browser.
How do I connect the MCP Inspector to my local server?
Start the inspector with `npx @modelcontextprotocol/inspector`, set Transport Type to `SSE` and URL to `http://localhost:8787/sse`, then click “Connect”. You will be prompted to log in (any email/password works in the mock flow).
How do I deploy the server to Cloudflare?
Create a KV namespace with `npx wrangler kv namespace create OAUTH_KV`, paste the returned ID into your `wrangler.jsonc` file, then run `npm run deploy`.
How do I connect Claude Desktop to a remote instance?
Update your Claude Desktop config file to use the `mcp-remote` command and point it to your `workers.dev` URL (e.g., `https://worker-name.account-name.workers.dev/sse`). Restart Claude; a browser window will open for login.
How can I debug connection issues?
Try connecting directly from the command line with `npx mcp-remote http://localhost:8787/sse`. In rare cases, clearing the `~/.mcp-auth` folder can resolve authentication problems. Restarting Claude may also help.
Basic information
More Cloud & Infrastructure MCP servers
container-use
aluzzardiDevelopment environments for coding agents. Enable multiple agents to work safely and independently with your preferred stack.
MCP Server that interacts with Azure AI Foundry (experimental)
azure-ai-foundryA MCP Server for Azure AI Foundry: it's now moved to cloud, check the new Foundry MCP Server
container-use
daggerDevelopment environments for coding agents. Enable multiple agents to work safely and independently with your preferred stack.
AWS MCP Servers
awslabsOpen source MCP Servers for AWS
Sample Serverless MCP Servers
aws-samplesSample implementations of AI Agents and MCP Servers running on AWS Serverless compute
Comments