Power your agent
การติดตั้งจากผู้เผยแพร่ ไม่ใช่เอนด์พอยต์ของ PlugMCP
URL แบบรีโมตหรือคำสั่งบนเครื่องของคุณ? การติดตั้งสองแบบต่างกันอย่างไร
รันบน PlugMCP
A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.
เปิดจากผู้เผยแพร่
This project contains two main components:
A Spring Boot application that functions as an MCP server. It makes tools available to AI models by converting OpenAPI specifications into MCP tools.
This project is designed to be run using Docker Compose, which orchestrates all the services (Java MCP Server, Python MCP Server, Sample Tools App, and MCP Client).
mcp-client requires a valid OpenAI API key. Before running, edit mcp-client/config.yml and replace YOUR_OPENAI_API_KEY with your actual key.To build all the Java projects and install Python dependencies, use the provided build script:
chmod +x build-all.sh
./build-all.sh
To start all services (Java MCP Server, Python MCP Server, Sample Tools App, and MCP Client) using Docker Compose:
docker-compose up --build
Once running, the services will be accessible on the following ports on your host machine:
http://localhost:18001http://localhost:18002http://localhost:18003http://localhost:18004The mcp-server acts as a tool provider for the mcp-client. The client connects to the server to get a list of available tools, which it then uses to interact with the configured LLM. This allows for a separation of concerns, where the server manages the tools and the client focuses on the interaction with the LLM.
Once you have a service with an OpenAPI specification, you can register it as a tool in either the Python or Java MCP server.
Add the tool to the tools list in the python_mcp_server/mcp_server.yml file. For example, to add the sample-tools-app:
mcp_server:
host: 0.0.0.0
port: 18002
tools:
- description: API for managing users and orders
name: sample-tools
rest_api_url: http://sample-tools-app:18003/v3/api-docs
The Java MCP Server can load tools on startup from an mcp_server.yml file located in its /config directory. An example is provided at mcp-server/config/mcp_server.yml.
Alternatively, you can dynamically register tools by sending a POST request to the /api/register-openapi endpoint. When running in Docker, use the service name (e.g., sample-tools-app) as the hostname for the source URL.
curl -X POST http://localhost:18001/api/register-openapi \
-H "Content-Type: application/json" \
-d '{
"type": "URL",
"source": "http://sample-tools-app:18003/v3/api-docs",
"friendlyName": "sample-tools",
"description": "Tools for managing users and orders"
}'
This project includes a full end-to-end test to verify that all components (mcp-server-java, sample-tools-app, and mcp-client) are working together correctly.
mcp-client requires a valid OpenAI API key to function. Before running the test, you must edit the mcp-client/config.yml file and replace the placeholder YOUR_OPENAI_API_KEY with your actual key.A shell script test-e2e.sh is provided to automate the entire process.
Make the script executable:
chmod +x test-e2e.sh
Run the script:
./test-e2e.sh
The script will:
mcp-client that requires it to use a tool from the sample-tools-app.Contributions are welcome! Please feel free to submit a pull request.
This project is licensed under the MIT License. See the LICENSE file for more information.
A Spring Boot application that functions as an MCP server. It makes tools available to AI models by converting OpenAPI specifications into MCP tools.
เครื่องมือ
ยังไม่มีรายการเครื่องมือ เซิร์ฟเวอร์แบบรีโมตจะได้รายการหลังการตรวจสอบสถานะครั้งแรก
FAQ
คัดลอกคอนฟิก stdio สำหรับ Cursor, Claude Desktop, Claude Code หรือ VS Code ได้ในหน้านี้ คอนฟิกนี้ใช้การติดตั้งของผู้เผยแพร่ ไม่ใช่พร็อกซีของ PlugMCP
ระเบียนนี้ไม่ได้ระบุตัวแปรสภาพแวดล้อมที่จำเป็น
Augmentia De ทำงานแบบโลคัลผ่าน stdio
ตรวจสอบล่าสุดเมื่อ 2026-11-05 A Spring Boot application that functions as an MCP server. It makes tools available to AI models by converting OpenAPI specifications into MCP tools.
