MCP Inventory Manager
Manages items and suppliers in a PostgreSQL database, providing CRUD operations, stock transfers, and supplier management through natural language commands.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
MCP Inventory Manager
An AI-powered inventory management system built with FastAPI, Model Context Protocol (MCP), and LangChain. A conversational AI agent (powered by Ollama/Llama 3.2) manages items and suppliers in a PostgreSQL database through natural language commands.
Academic project for the Enterprise Application Integration (IS) course — Master's in Computer Engineering, University of Coimbra, 2025/2026.
Features
Natural Language Interface — manage inventory through a chat UI powered by an LLM agent
Full CRUD — create, read, update, and delete items and suppliers
Stock Transfers — transfer quantities between items with validation
Supplier Management — link items to suppliers, lookup by name
MCP Server — tools exposed via the Model Context Protocol for AI agent integration
REST API — standard FastAPI endpoints alongside the AI chat interface
Related MCP server: Stock MCP
Architecture
Browser (Chat UI)
│
│ HTTP
▼
FastAPI Server
│
├── /chat endpoint ──> LangChain Agent (Ollama/Llama 3.2)
│ │
│ MCP Tools (stdio)
│ │
│ MCP Server (FastMCP)
│ │
├── REST endpoints ──────────┤
│ │
▼ ▼
SQLModel / PostgreSQLThe LangChain agent uses a ReAct pattern with MCP tools to interpret user requests, call the appropriate inventory operations, and return natural language responses.
Tech Stack
Component | Technology |
Language | Python 3.12 |
Web Framework | FastAPI + Uvicorn |
AI Agent | LangChain + LangGraph |
LLM | Ollama (Llama 3.2) |
MCP | FastMCP (Model Context Protocol) |
ORM | SQLModel |
Database | PostgreSQL |
Package Manager | uv |
Getting Started
Prerequisites
Setup
# Install dependencies
uv sync
# Configure database connection
# Create a .env file with:
DATABASE_URL="postgresql://postgres:postgres@127.0.0.1:5432/mcp_is_project"
# Run the server
uv run python main.pyThe app will be available at:
Chat UI:
http://localhost:8000/uiREST API:
http://localhost:8000/docs
Project Structure
MCP-IS-PROJECT/
├── main.py # FastAPI app with REST endpoints and chat
├── mcp_server.py # MCP server with all inventory tools
├── agent.py # LangChain ReAct agent with Ollama
├── models.py # SQLModel data models (Item, Supplier)
├── services.py # Business logic layer
├── database.py # Database connection and setup
├── static/ # Chat UI frontend
├── db_model.txt # Database schema documentation
└── pyproject.toml # Dependencies and project configTeam
Francisco Pereira
Tiago Mendes
License
This project is licensed under the MIT License — see the LICENSE file for details.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
AI inventory management: stock, Demand ESP reorder recommendations and purchase orders.
MCP server unifying ERPs, CRMs, APIs and knowledge base for Claude, ChatGPT and Gemini.
Hosted MCP for e-commerce: live product catalog, stock, and pricing for AI agents.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceInventory Management AI - MCP server providing AI-powered tools and automation by MEOK AI Labs8 npm36 PyPIMIT
- AlicenseNot gradedqualityBmaintenanceUnifies inventory data from four sources (Jikeyun, Supor factory, WeChat Excel, RPA) with timestamps, and exposes MCP tools for AI agents to query stock levels.14 npmISC
- AlicenseNot gradedqualityDmaintenancePredictive supply-chain MCP server that forecasts material confirmation risks and enables AI clients to interact with the system via natural language.MIT
- FlicenseNot gradedqualityBmaintenanceEnables AI-powered warehouse inventory management through MCP, with tools to search products, check stock levels, receive and issue stock, and add products under human-in-the-loop confirmation.-