MCP Test Project
A simple MCP server/client implementation for testing Hermes Agent, exposing basic tools like echo, add, and get_server_info via stdio.
README
MCP Test Project
A simple Model Context Protocol (MCP) server/client implementation for testing Hermes Agent with GitHub MCP.
Overview
This project demonstrates a basic MCP (Model Context Protocol) implementation with:
- MCP Server (
src/mcp_test/server.py) - Exposes tools via stdio transport - MCP Client (
src/mcp_test/client.py) - Connects to and tests MCP servers - Data Models (
src/mcp_test/models.py) - Pydantic models for tool calls and results - Tests (
tests/) - Unit tests for models and server functionality
Project Structure
mcp/
├── pyproject.toml # Project configuration (dependencies, build, tools)
├── Makefile # Development commands
├── .gitignore # Git ignore rules
├── README.md # This file
├── src/
│ └── mcp_test/
│ ├── __init__.py # Package exports
│ ├── models.py # Pydantic data models
│ ├── server.py # MCP server implementation
│ └── client.py # MCP client implementation
└── tests/
├── __init__.py # Test package
├── test_models.py # Tests for data models
└── test_server.py # Tests for server functionality
Requirements
- Python 3.10+
- Dependencies listed in
pyproject.toml
Installation
# Install in development mode with dev dependencies
make install
# Or manually:
pip install -e ".[dev]"
Running the Server
# Using make
make run-server
# Or directly
python -m mcp_test.server
The server runs over stdio and exposes these tools:
echo- Echo back a messageadd- Add two numbersget_server_info- Get server metadata
Running the Client
# Using make
make run-client
# Or directly
python -m mcp_test.client
The client will connect to the server (started as a subprocess) and run test calls.
Running Tests
# Run all tests with coverage
make test
# Or directly
pytest -v
Code Quality
# Run linting
make lint
# Format code
make format
# Type checking
make typecheck
# All checks
make check
Development
Adding a New Tool
- Add the tool definition in
server.pyin thelist_tools()handler - Add the implementation method (e.g.,
_tool_your_tool) - Add the tool call handler in
call_tool() - Add a convenience method in
client.pyif needed - Write tests in
tests/test_server.py
Running a Single Test
pytest tests/test_models.py::test_tool_call_creation -v
License
MIT License - see LICENSE file for details.
Related
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。