home / mcp / mcp curl mcp server
An MCP Server wrapper around the curl CLI utility
Configuration
View docs{
"mcpServers": {
"247arjun-mcp-curl": {
"command": "mcp-curl",
"args": [],
"env": {
"DEBUG": "true"
}
}
}
}You have a dedicated MCP server that lets AI assistants perform HTTP requests using curl-like capabilities directly from their environment. It supports GET, POST, PUT, DELETE, file downloads, custom headers, timeouts, redirects, and JSON data handling, with safe subprocess execution and input validation to keep things secure.
You can run the MCP server locally and connect your MCP client to it, either by invoking the server directly or via a higher-level client that routes requests through the MCP protocol. Use the built-in curl tools to perform common HTTP actions from within your AI workflows. Start with a simple GET to fetch data, then move to POST/PUT for sending JSON payloads, and use the download tool to save files when needed. The server handles custom headers, timeouts, redirects, and a configurable User-Agent string to tailor requests to your environment.
Prerequisites: you need Node.js (version 18 or newer) and npm. You also need curl installed on your system for compatibility in some environments.
Option A — Install globally with npm (recommended) 1. Install globally 2. Verify the installation 3. Start using the MCP server.
# Install globally
npm install -g @247arjun/mcp-curl
# Or install locally in your project
npm install @247arjun/mcp-curlOption B — Install from source 1. Clone the repository 2. Install dependencies 3. Build the project 4. Optionally link globally.
# Clone the repository
git clone https://github.com/247arjun/mcp-curl.git
cd mcp-curl
# Install dependencies
npm install
# Build the project
npm run build
# Optional: Link globally
npm linkOption C — Install directly from GitHub 1. Install from GitHub 2. Use the package as a global command.
# Install directly from GitHub
npm install -g git+https://github.com/247arjun/mcp-curl.gitThis MCP server can be wired into your Claude Desktop environment so you can issue HTTP requests programmatically from your assistant. The following configurations show how to expose the server as a local stdio process.
{
"mcpServers": {
"mcp-curl": {
"command": "mcp-curl",
"args": []
}
}
}Alternative: use npx to avoid a global install. The MCP server can be launched via npx with the package name.
{
"mcpServers": {
"mcp-curl": {
"command": "npx",
"args": ["@247arjun/mcp-curl"]
}
}
}Local development setup can point to a built index. Provide the absolute path to the built index file.
{
"mcpServers": {
"mcp-curl": {
"command": "node",
"args": ["/absolute/path/to/mcp-curl/build/index.js"]
}
}
}Input validation and sanitization for all curl arguments are performed. Subprocess execution is safe, using spawn rather than shell. No arbitrary shell commands are executed. Input validation relies on schema validation to prevent malformed requests.
Note before using: you can perform HTTP requests using dedicated tools that support common HTTP methods and options.
1) curl_get: Make HTTP GET requests with optional headers, timeout, redirects, and a custom User-Agent.
2) curl_post: Make HTTP POST requests with JSON data or raw body data, plus headers and content type.
3) curl_put: Make HTTP PUT requests with JSON data or raw body data and headers.
4) curl_delete: Make HTTP DELETE requests with optional headers.
5) curl_download: Download files with support for resuming and redirects.
6) curl_advanced: Execute curl with a custom set of arguments for advanced scenarios.
Prerequisites: Node.js 18.0.0 or higher, npm, and curl. Building: run the build script. Testing: run the test suite. Linting: run lint commands.
npm run buildTo test the server manually, run the built server and verify it starts as expected.
Common issues include missing command, permissions, and MCP server not appearing in the client. Ensure the correct command path is used, verify execute permissions, and restart the client after updating the configuration.
Debugging can be done by enabling verbose logging and sending test inputs to the server.
The MCP server is designed for local development and integration with client environments that support the MCP protocol. Follow your environment’s best practices for process management and logging when deploying.
The project includes source, built assets, tests, and example usage files to help you get started and validate functionality.
Verify that the server is working by running the built index and observing the startup message indicating the server is running in its chosen mode.
Make HTTP GET requests with optional headers, timeout, redirects, and a custom User-Agent string.
Make HTTP POST requests with JSON data or raw body data, plus headers and content type.
Make HTTP PUT requests with JSON data or raw body data and headers.
Make HTTP DELETE requests with optional headers.
Download files with support for resuming and redirects.
Execute curl with a custom set of arguments for advanced scenarios.