ShopBot exposes order-status lookup and knowledge-base answers through a local Python MCP server backed by a store database and RAG pipeline.
ShopBot exposes order-status lookup and knowledge-base answers through a local Python MCP server backed by a store database and RAG pipeline.
The server exposes two implemented tools: check_order_status and search_knowledge_base. Order lookup reads product, status, order time and tracking information from a configured SQL database. The knowledge tool uses text files from the repository's knowledge_base directory. It does not implement placing orders, cancelling purchases or issuing refunds.
DB_CONNECTION_STRING configures the SQLAlchemy database connection. The RAG implementation uses Google's embedding and chat integrations, not an OpenAI pipeline. It builds or loads a local FAISS index from the knowledge files. The project uses SDK 1 FastMCP imports, so install compatible dependencies or migrate those imports before launching the local stdio script. Use only a FAISS index you created or trust because the source enables deserialization when loading it.
Start with a test order in a database you control. Compare the returned tracking and status fields with the row, then ask a policy question answered explicitly by a knowledge file. Check the retrieved answer against that file. Database access and model access need their own credentials; an MCP connection does not prove that either backing service is configured or that an answer matches current store policy.
Check the linked publisher documentation and your connected client's tool list. Names and availability can vary with the server version and account permissions.
| Tool | What it does |
|---|---|
| check_order_status | Read status and tracking for an order. |
| search_knowledge_base | Answer a question using the configured knowledge base. |
{
"mcpServers": {
"shopbot": {
"command": "python",
"args": [
"/ABSOLUTE_PATH/shopbot-ai/mcp/shopbot_server.py"
],
"env": {
"DB_CONNECTION_STRING": "YOUR_DATABASE_CONNECTION_STRING"
}
}
}
}The project Python dependencies, compatible MCP SDK 1 imports, an orders database, Google model credentials and prepared knowledge-base files.
| Variable | Description | Required |
|---|---|---|
| DB_CONNECTION_STRING | SQLAlchemy connection string for the store database. | Yes |
# 🤖 AI Customer Support Bot - MCP Server <div align="center"> 
# Airtable Airtable is the database and operations layer for your agents — whether running product, marketing, sales, ops, HR, or a custom business
Ask a business's official customer support and get their verified answer, with sources.
BackEngine is a multi-tenant SaaS platform that ingests all of an organization's customer and prospect communications — Slack threads, emails, call and meeting transcripts, and support tickets — and…
Customer support for MCP servers: work your tickets from Claude or any MCP client. Service key auth.
Turn a support ticket into a structured case brief, or score its L2 escalation readiness.