ryanjoachim/mcp-batchit

将多个MCP工具调用批量处理成单一“batch_execute”请求——减少AI代理的开销和令牌使用。

---

目录

1. 引言
2. 为什么使用 BatchIt
3. 主要特征与局限性
4. 安装与启动
5. 多相应用

• 实施阶段
• 信息收集
• 仅LLM步骤（列表代码定义）
• 文档创建

6. 常见问题解答
7. 许可

---

引言

⚠️ 通知：工作进行中

该项目正在积极开发，以应对若干复杂挑战：

• 保持与现有MCP服务器的向后兼容性
• 解决多连接客户端（Cline、Roo、Claude Desktop）的传输复杂性
• 创建适合初学者的实现

虽然功能正常，但随着解决方案的完善，请期待持续改进和变化。

MCP BatchIt 是模型上下文协议（MCP）生态系统中的一个简单聚合服务器。它只暴露一个工具：batch_execute。与其在单独消息中调用多个MCP工具（如fetch]、read_file、[[T3]、write_file等）不同，你可以将它们批处理**在同一个聚合请求中。

---

Table of Contents

1. Introduction
2. Why Use BatchIt
3. Key Features & Limitations
4. Installation & Startup
5. Multi-Phase Usage
   • Implementation Phases
     • Information Gathering
     • LLM‐Only Step (List Code Definitions)
     • Document Creation
6. FAQ
7. License

---

Introduction

⚠️ NOTICE: Work in Progress

This project is actively being developed to address several complex challenges:

• Maintaining backwards compatibility with existing MCP servers
• Resolving transport complexities with multi-connection clients (Cline, Roo, Claude Desktop)
• Creating a beginner-friendly implementation

While functional, expect ongoing improvements and changes as we refine the solution.

MCP BatchIt is a simple aggregator server in the Model Context Protocol (MCP) ecosystem. It exposes just one tool: batch_execute. Rather than calling multiple MCP tools (like fetch, read_file, create_directory, write_file, etc.) in separate messages, you can batch them together in one aggregator request.

This dramatically reduces token usage, network overhead, and repeated context in your AI agent or LLM conversation.

---

Why Use BatchIt

• One Action per Message Problem: Normally, an LLM or AI agent can only call a single MCP tool at a time, forcing multiple calls for multi-step tasks.

• Excessive Round Trips: 10 separate file operations might require 10 messages → 10 responses.

• BatchIt’s Approach:

  1. Takes a single batch_execute request.
  2. Spawns (or connects to) the actual target MCP server (like a filesystem server) behind the scenes.
  3. Runs each sub-operation (tool call) in parallel up to maxConcurrent.
  4. If one sub-op fails and stopOnError is true, it halts new sub-ops.
  5. Returns one consolidated JSON result.

---

Key Features & Limitations

Features

1. Single “Batch Execute” Tool

   • You simply specify a list of sub‐ops referencing your existing MCP server’s tools.

2. Parallel Execution

   • Run multiple sub-ops at once, controlled by maxConcurrent.

3. Timeout & Stop on Error

   • Each sub-op races a timeoutMs, and you can skip remaining ops if one fails.

4. Connection Caching

   • Reuses the same connection to the downstream MCP server for repeated calls, closing after an idle timeout.

Limitations

1. No Data Passing Mid-Batch