Intelligent AI Order Support Policy Assistant
A Spring Boot application for order-support workflows that combines order data, policy retrieval, remote MCP tools, and AI-powered guidance for support agents.
接入你的智能体
從發布者安裝。不是 PlugMCP 端點。
遠端 URL 還是本機指令? 兩種安裝方式的差異
來自發布者
A Spring Boot application for order-support workflows that combines order data, policy retrieval, remote MCP tools, and AI-powered guidance for support agents.
This is is AI Chatbot which uses RAG and MCP Server to connect internal database to provide intelligent Order Support and provides assistance related to order management policies. It is to augment DevOPS team.
The assistant supports a real business workflow:
See docs/architecture.md for the detailed system, container, and component diagrams.
The app reads the following configuration values:
PowerShell example:
$env:MONGODB_URI="mongodb://localhost:27017/orderdb"
$env:REDIS_HOST="localhost"
$env:REDIS_PORT="6379"
$env:OPENAI_API_KEY="your-openai-api-key"
$env:MCP_API_KEY="your-internal-mcp-key"
$env:MCP_INQUIRY_BASE_URL="http://localhost:8081"
$env:MCP_PROCESSING_BASE_URL="http://localhost:8082"
$env:COPILOT_POLICY_AGENT_BASE_URL="https://copilot.example.com"
$env:COPILOT_POLICY_AGENT_API_KEY="your-copilot-key"
mvn spring-boot:run
The app runs at:
The browser UI calls the support API when a user enters a question.
POST /api/support/query
Content-Type: application/json
Request body example:
{
"customerId": "C-991",
"orderId": "O-1001",
"question": "My order is delayed. Can I get a shipping refund?",
"confirmCancellation": false
}
Response example:
{
"answer": "Based on the order status and applicable shipping policy, this shipment is eligible for a partial review..."
}
Invoke-RestMethod -Method Post -Uri "http://localhost:8080/api/support/policies/ingest"
Invoke-RestMethod -Method Post -Uri "http://localhost:8080/api/support/orders/seed"
$body = @{
customerId = "C-991"
orderId = "O-1001"
question = "My order is delayed. Can I get shipping fee refund or compensation?"
confirmCancellation = $false
} | ConvertTo-Json
Invoke-RestMethod -Method Post `
-Uri "http://localhost:8080/api/support/query" `
-ContentType "application/json" `
-Body $body
Invoke-RestMethod -Method Post `
-Uri "http://localhost:8080/api/support/chat/end?userId=C-991"
The application is designed around remote services rather than local in-process tool calls:
Live interaction state is kept in Redis while the session is active. When the user ends the chat, the completed conversation is moved to MongoDB for durable archival and later review.
The platform enforces a Redis-backed token budget per user:
Before the model call, the service compacts the flow into a smaller, optimized history:
This reduces token waste while preserving the decision context needed by the model.
The project includes starter policy documents:
Default seeded records:
This is is AI Chatbot which uses RAG and MCP Server to connect internal database to provide intelligent Order Support and provides assistance related to order management policies. It is to augment Dev
總覽
This is is AI Chatbot which uses RAG and MCP Server to connect internal database to provide intelligent Order Support and provides assistance related to order management policies. It is to augment DevOPS team.
工具
尚無工具清單。遠端伺服器會在首次健康檢查後取得工具清單。
常見問題
在本頁複製適用於 Cursor、Claude Desktop、Claude Code 或 VS Code 的 stdio 設定。它使用發布者的安裝方式,而不是 PlugMCP 代理伺服器。
這筆記錄未列出必要的環境變數。
IntelligentAIOrderSupportPolicyAssistant 透過 stdio 在本機執行。
上次檢查時間:2026-10-28。 This is is AI Chatbot which uses RAG and MCP Server to connect internal database to provide intelligent Order Support and provides assistance related to order management policies. It is to augment DevOPS team.