Skip to main content
Glama
Sign In

Dynamic API Server (MCP)

by 1049861657
GitHub
TypeScript
OverviewInspectNewSchemaRelated ServersReviewsScore
Need Help?View Source CodeReport Issue

动态API服务器

一个基于MCP协议的模块化动态API服务器，为AI助手提供丰富的工具能力，同时显著减少prompt token消耗。

特点

  • 模块化API设计：每个API都是独立模块，易于维护和扩展
  • 动态发现与加载：自动扫描和加载API目录中的模块
  • 元工具设计：AI仅需知道三个元工具，而不是全部API
  • 热更新支持：监控API文件变化，无需重启即可更新
  • 类型安全：完整的TypeScript类型定义
  • 丰富的文档：每个API都包含详细说明和使用示例

工作原理

本项目解决了传统MCP工具实现中API数量增长导致prompt负担过重的问题。通过元工具设计模式，将所有具体API实现从prompt中剥离，仅保留三个核心元工具：

  1. listAllApis：列出所有可用API，可按分类筛选
  2. getApiDetails：获取特定API的详细信息和使用方法
  3. executeApi：执行指定的API，接收API ID和参数

AI可以先探索可用API，获取详情，然后再执行，而不需要预先了解所有API的细节。

目录结构

/MCP /src # 源代码目录 /core # 核心模块 registry.ts # API注册中心 loader.ts # 动态加载机制 server.ts # MCP服务器核心 /apis # API模块目录 /config # 配置文件目录 api-config.ts # API全局配置 /utils # 实用工具库 index.ts # 入口文件

构建和运行

# 编译并启动 npm run start:dev

添加新API

创建新API非常简单，只需要在src/apis目录下的适当子目录中创建新文件：

// src/apis/your-category/your-api.ts import { z } from 'zod'; import { ApiCategories } from '../../core/registry'; export default { id: "yourApiId", name: "你的API名称", description: "API功能描述", category: ApiCategories.COMMON, // 选择适当的分类 schema: { // 使用Zod定义参数 param1: z.string().describe("参数1说明"), param2: z.number().optional().describe("可选参数2说明") }, handler: async ({ param1, param2 = 0 }) => { try { // 实现API逻辑 return `你的API结果: ${param1}, ${param2}`; } catch (error) { throw new Error(`执行失败: ${error.message}`); } }, examples: [ { description: "使用示例1", params: { param1: "示例值", param2: 42 } } ] };

保存文件后，服务器会自动发现并加载新API，无需重启。

维护和扩展

  • 添加新分类：在src/core/registry.ts中的ApiCategories对象添加新分类
  • 修改API行为：直接编辑相应的API模块文件
  • 修改加载行为：调整src/core/loader.ts中的配置或行为

许可证

MIT

This server cannot be installed

-
security - not tested
-
license - not tested
-
quality - not tested

hybrid server

The server is able to function both locally and remotely, depending on the configuration or use case.

A modular dynamic API server based on the MCP protocol that provides rich tool capabilities for AI assistants while significantly reducing prompt token consumption.

  1. 特点
    1. 工作原理
      1. 目录结构
        1. 构建和运行
      2. 添加新API
        1. 维护和扩展
          1. 许可证

            Related MCP Servers

            • MCP BatchIt

              ryanjoachim
              -
              security
              F
              license
              -
              quality
              A simple aggregator server that allows batching multiple MCP tool calls into a single request, reducing token usage and network overhead for AI agents.
              Last updated -
              14
              JavaScript

            • human-mcp

              upamune
              -
              security
              -
              license
              -
              quality
              An MCP server that allows AI assistants to utilize human capabilities by sending requests to humans and receiving their responses through a Streamlit UI.
              Last updated -
              23
              Python
              MIT License

            • ServiceTitan MCP Server

              JordanDalton
              A
              security
              F
              license
              A
              quality
              An MCP server that allows AI assistants to interact with the ServiceTitan API, requiring client credentials for authentication.
              Last updated -
              454
              TypeScript

            • OpenAPI to MCP Server

              TykTechnologies
              -
              security
              A
              license
              -
              quality
              A tool that creates MCP (Model Context Protocol) servers from OpenAPI/Swagger specifications, enabling AI assistants to interact with your APIs.
              Last updated -
              47
              9
              TypeScript
              MIT License
              • Apple

            View all related MCP servers

            Appeared in Searches

            New MCP Servers

            MCP directory API

            We provide all the information about MCP servers via our MCP API.

            curl -X GET 'https://glama.ai/api/mcp/v1/servers/1049861657/MCP'

            If you have feedback or need assistance with the MCP directory API, please join our Discord server