What is Mcp Execution Engine

The mcp-execution-engine is a modular and lightweight execution engine designed for AI agents, automation pipelines, and multi-step tool workflows. It allows for dynamic execution of multiple tools in sequence and supports various interaction methods including CLI, RESTful API, and Telegram Bot, all within a single Docker image.

Use cases

Use cases for the mcp-execution-engine include automating workflows for AI agents, integrating multiple tools in a sequence for complex tasks, and providing a unified interface for various interaction methods in automation pipelines.

How to use

To use the mcp-execution-engine, first install Docker, then pull the Docker image using ‘docker pull sidwoong/xxx-mcp-execution-engine’. You can start it in CLI mode or API server mode by running the appropriate Docker commands with your environment configuration.

Key features

Key features of the mcp-execution-engine include dynamic execution of tool chains, interaction via CLI, HTTP API, or Telegram, support for custom tools through a modular system, and Docker-native deployment with no local dependencies.

Where to use

The mcp-execution-engine is suitable for backend execution engines, plugin-based AI systems, and hybrid AI workflows that require both on-chain and off-chain processing.

🧠 XXX MCP Execution Engine

A modular, lightweight ToolCall execution engine designed for AI agents, automation pipelines, and multi-step tool workflows. It supports CLI interaction, RESTful API service, and Telegram Bot access — all packaged in a single Docker image.

✨ Product Overview

XXX MCP Execution Engine provides a unified runtime for structured ToolCall chains. It allows LLM agents or users to:

• 🔧 Dynamically execute multiple tools in sequence
• 💬 Interact via terminal, HTTP API, or Telegram
• 📦 Plug in custom tools via the modular tools/ system
• 🧱 Use anywhere via Docker — no manual setup required

This project is inspired by the idea of decoupling agent planning from tool execution, making it perfect for backend execution engines, plugin-based AI systems, or on-chain/off-chain hybrid AI workflows.

✅ Built-in modes: CLI / API Server / Telegram Bot
🐳 Docker-native, zero local dependency
🔌 Supports future extensibility with user-defined tools

🐳 How to Install Docker (for Beginners)

macOS / Windows / Linux

1. Download from the official Docker site:
   👉 https://www.docker.com/products/docker-desktop

2. Follow the installation steps.

3. After installation, run:

docker --version

✅ If you see version output, Docker is installed.

⚙️ Environment Configuration (.env)

⚙️ Environment Configuration (.env)

MCP_LANG=ZH
OPENAI_API_KEY=sk-xxxxxxxxxx
WALLET_ADDRESS=0xxxxxxxxxxx

🚀 Quick Start

Pull the Docker Image

docker pull sidwoong/xxx-mcp-execution-engine

Start CLI Mode

docker run --rm -it --env-file .env sidwoong/xxx-mcp-execution-engine start cli

API Server Mode

docker run --rm -it --env-file .env -p 8080:8080 sidwoong/xxx-mcp-execution-engine start server

Telegram Bot Mode

docker run --rm -it --env-file .env sidwoong/xxx-mcp-execution-engine start telegram-bot

🧩 Features

🔍 Preview

🧠 CLI Mode (start cli)

Launch interactive execution in terminal.

Interactively execute ToolCall chains step-by-step.
Each tool call is clearly displayed with execution flow, tool name, and results.

🌐 API Server Mode (start server)

Run as an HTTP API service using FastAPI.

🤖 Telegram Mode (start telegram-bot)

Turn your bot into a private assistant.

Easily turn MCP into your personal assistant inside Telegram.
Users can chat naturally and trigger multi-step ToolCall executions.

📮 Contact

Author: Sid Woong
GitHub: SidWoong
X (Twitter): @SidWoong

✅ Project is actively maintained. Content continuously updated. For any questions, feel free to contact author on X: @SidWoong