HTTP MCP Server

A Model Context Protocol (MCP) server that enables AI agents to send HTTP requests to any endpoint with full control over methods, headers, query parameters, and request bodies.

Note: This MCP server only supports stdio transport.

MCP Badge from LobeHub

Table of Contents

• Features
• Installation
• Configuration
• Local Development
  • Install
  • Build
  • Run

Features

• Support for all major HTTP methods (GET, POST, PUT, PATCH, DELETE)
• Custom headers configuration
• Query parameter handling with proper URL encoding
• Request body support for POST/PUT/PATCH
• Comprehensive error handling
• TypeScript with strict type safety
• Clean, modular architecture
• STDIO transport support for seamless integration with MCP clients

Installation

Install the package globally using npm:

npm install -g @jules-tnk/ts-http-mcp

Or use npx to run it directly:

npx @jules-tnk/ts-http-mcp

Configuration

Claude Desktop

Add the following configuration to your Claude Desktop config file (claude_desktop_config.json):

Windows: %APPDATA%\Claude\claude_desktop_config.json macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Linux: ~/.config/claude/claude_desktop_config.json

{
	"mcpServers": {
		"http-mcp": {
			"command": "npx",
			"args": ["@jules-tnk/ts-http-mcp"]
		}
	}
}

After updating the configuration, restart Claude Desktop to load the MCP server.

Other MCP Clients

For other MCP clients that support stdio transport, configure them to run:

npx @jules-tnk/ts-http-mcp

Local Development

Install

pnpm install

Build

pnpm build

Run

pnpm start

Testing

To test the MCP server locally, you can run it directly and interact with it through stdio:

pnpm start

The server will start and listen for MCP protocol messages on stdin/stdout.

Usage

Once configured with your MCP client, the HTTP MCP server provides the following capabilities:

• GET requests: Retrieve data from any HTTP endpoint
• POST requests: Send data to endpoints that accept POST requests
• PUT/PATCH requests: Update resources on remote servers
• DELETE requests: Remove resources from remote servers
• Custom headers: Add authentication, content-type, and other headers
• Query parameters: Include URL parameters with proper encoding
• Request bodies: Send JSON, form data, or raw content

The server handles all the HTTP communication details, error handling, and response processing, making it easy for AI agents to interact with web APIs and services.

License

MIT License - see the LICENSE.md file for details.