MCP-ANY-REST

Provides an MCP runtime to connect RESTful services to AI via MCP with transport options and a structured config.

rust

2

GitHub Stars

rust

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "xiangweizeng-mcp-any-rest": {
      "command": "cargo",
      "args": [
        "run",
        "--bin",
        "mcp-any-rest",
        "--",
        "--transport",
        "stdio"
      ]
    }
  }
}

You can run MCP-ANY-REST to connect REST services to AI using the Model Context Protocol. It lets you configure how your REST APIs are accessed, orchestrated, and exposed to AI clients in a consistent, zero‑code way. This server focuses on safe, efficient integration so your LLMs can safely query and act on business data via REST APIs.

How to use

Use an MCP client to connect to the MCP-ANY-REST server. You will choose a transport (http for remote access, or stdio for local/in-process usage) and point the client at the server with a proper configuration directory. You can run the server with a specific configuration directory or rely on a default config location relative to the executable.

How to install

Prerequisites: Install Rust 1.70+ (2021 edition). You also need a REST service that you want to access via MCP-ANY-REST and a supported environment to run the server.

Step 1: Clone the project and enter the directory.

Step 2: Build the project. You have two options: use the provided build scripts or build directly with Cargo.

# Build using provided scripts (Windows)
.\build_and_copy.bat

# Build using provided scripts (Linux/macOS)
./build_and_copy.sh

# Build directly with Cargo (release)
cargo build --release

Additional configuration and startup notes

Configuration files live in a config directory. If you specify a custom directory, pass it to the server; otherwise, the server automatically uses a config directory located relative to the executable.

Example start commands show how to run the server with a specific transport and config directory.

Troubleshooting and tips

If you want to run in http mode with a custom config, use a command that points to your config directory and selects the http transport.

If you prefer local development or in-process usage, run the server with stdio transport once you have a compatible config in your local directory.

Notes on configuration layout

Your configuration directory should contain core files such as config.json, modules.json, and a presets folder with preconfigured module sets. This layout enables the server to discover and load modules at runtime without code changes.

Available tools

zml_cli

Unified ZML command line tool to list modules and compile ZML to JSON.

service_composer

Service composition layer that provides a unified interface to orchestrate modules and services.

zml_compiler

Compiler that translates ZML module definitions into JSON configuration used at runtime.

mcp_runtime

Runtime core implementing MCP protocol interactions and RMCP compliance.