mcpField
M
MCP客户端应用

MCP客户端应用

Local
github.com/WilliamSuiselfAI & Agents1Updated 82w ago

About MCP客户端应用

MCPHOLDER is a mobile MCP client for connecting to services hosted on Cloudflare. It combines model-based chat with pages for managing services, testing connections, and authorizing a service before its tools are used in chat.

Key features

  1. 01Chat with selectable models
  2. 02View available MCP services and status
  3. 03Add and test service connections
  4. 04Authorize services through OAuth

Use cases

  1. 01Chat with a model on a phone
  2. 02Configure and check a Cloudflare MCP service
  3. 03Call connected MCP tools from an AI chat

MCP support

As an MCP client, it connects to Cloudflare-hosted MCP services and supports connection testing and OAuth authorization. It can discover service tools for use in AI conversations.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

MCP客户端应用

这是一个为AI大模型服务的MCP(model context protocol)客户端程序-MCPHOLDER,在手机上也能实现类似于claude desktop的功能。它可以将cloudflare上的云端服务汇集到一起,为AI大模型提供外部的支持。

功能

  • AI对话页面:用户可以选择不同的大模型进行对话,默认使用Claude-3 Sonnet 3.7
  • MCP服务页面:显示当前可用的MCP服务及其状态,并提供添加新服务的入口
  • MCP服务测试页面:测试新的MCP服务连接,输入服务名称、URL、端口等信息

技术栈

  • Flutter 3.x
  • Dart 3.x
  • font_awesome_flutter:用于图标展示
  • google_fonts:用于字体样式

快速开始

1、请修改lib/config目录下的openRouterApiKey,当前程序只支持openrouter的claude sonnet 3.7 2、由于网络环境问题,我只能上传flutter的lib库,因此以下安装方式不一定可用:

  1. 确保已安装Flutter环境
  2. 克隆此仓库
  3. 运行以下命令安装依赖:
flutter pub get
  1. 运行应用:
flutter run

项目结构

lib/
  ├── main.dart              # 应用入口
  ├── screens/               # 页面
  │   ├── chat_screen.dart   # AI对话页面
  │   ├── services_screen.dart  # MCP服务页面
  │   └── test_service_screen.dart  # MCP服务测试页面
  ├── widgets/               # 可复用组件
  ├── models/                # 数据模型
  └── utils/                 # 工具函数

MCP协议

Model Context Protocol (MCP) 是一种协议,允许AI模型通过外部工具和API执行操作。这使得AI助手能够执行实际任务,如发送电子邮件、查询数据或控制应用程序。

更多关于MCP协议的信息,请访问 Cloudflare的MCP文档。

演示

https://youtube.com/shorts/npAoVR9HOHM?feature=share

在测试服务页面,我们先: 填入服务名称:add, 再填入cloudflare的服务URL:https://remote-mcp.yunhaisui.workers.dev/sse 端口不用填写 API秘钥随便输入,但不能为空

下面点击注册客户端获取ID 成功后点击测试连接 然后会弹出一个认证的OAUTH授权按钮,点击启动oauth授权 在新弹出的webview页面点击approve 然后App会自动识别MCP中有几个tools

这时点击MCP服务,就可以看到新增了一个加法工具

在AI对话页面中输入一个简单的加法问题,claude sonnet会自动调用mcp服务,我使用的是openrouter服务,似乎有点慢,但最终也能得到正确的结果。下周我再到cloudflare上做几个好玩的MCP服务测试一下。

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.