Server Curl

What is Server Curl? Server Curl (@mcp-get-community/server-curl) is an MCP server that provides a curl-like tool allowing LLMs to make HTTP requests to any URL. It exposes a single tool named “curl” that returns full response information (status, headers, body). Note: the project/registry is archived and no longer maintained.

How to use Server Curl?

• Install: npx @michaellatman/mcp-get@latest install @mcp-get-community/server-curl
• Configure in your mcpServers manifest to run the server (example in the repo). The server runs in Node (via npx) and exposes the “curl” tool.
• Invoke the “curl” tool with parameters: url (required), method (optional, default GET), headers (optional object), body (optional string), timeout (optional ms, default 30000, max 300000). The tool returns status, statusText, headers, and body.

Key features

• Make HTTP requests to any URL
• Support for common HTTP methods: GET, POST, PUT, DELETE, PATCH, HEAD, OPTIONS
• Custom headers and request body support
• Configurable timeout (default 30000 ms, max 300000 ms)
• Full response details (status, headers, body)
• Easy install/run via npx and MCP manifest

Use cases

1. Allow LLM agents to fetch external APIs and data during workflows
2. Post data to web services or trigger webhooks from LLM-driven automations
3. Test and debug HTTP endpoints from within an MCP environment
4. Proxy or relay requests when an LLM needs external content
5. Integrate third-party APIs into model-based tools and pipelines

FAQ

• Is this project maintained?

No. The mcp-get community server registry is archived and no longer actively maintained.

• How do I install it?

Run: npx @michaellatman/mcp-get@latest install @mcp-get-community/server-curl

• Which HTTP methods are supported?

GET, POST, PUT, DELETE, PATCH, HEAD, OPTIONS.

• What are the timeout defaults/limits?

Default timeout is 30000 ms; maximum allowed is 300000 ms.

• What does the response include?

The response includes status, statusText, headers, and body.