Model-Context-Protocol (MCP) server that lets AI assistants securely read & write data in an Odoo ERP instance via standardized MCP resources and tools.
https://github.com/ivnvxd/mcp-server-odooStop switching between your AI assistant and Odoo's interface. This MCP server connects Claude, Cursor, and other AI tools directly to your Odoo instance, letting you query customers, create orders, and manage your ERP through natural conversation.
If you're already using Odoo for business operations, you know the drill: context switching between your AI assistant for analysis and Odoo's web interface for data access kills productivity. You end up copy-pasting data, manually cross-referencing records, or writing one-off scripts just to get simple answers.
This MCP server eliminates that friction entirely. Your AI assistant gets secure, real-time access to your Odoo data with full CRUD capabilities.
Business Intelligence Through Conversation:
Data Management Without Web Forms:
System Analysis and Reporting:
Secure by Design: Uses Odoo's built-in API key system and respects all existing access controls and record rules. No bypassing your security model.
Zero Context Switching: Query and modify ERP data without leaving your development environment or breaking conversation flow with your AI assistant.
Multiple Transport Options: Works with stdio for desktop apps like Claude Desktop, or HTTP for remote/server deployments. Pick what fits your infrastructure.
Smart Field Selection: Automatically returns relevant fields for AI processing, or specify exactly what you need. No more dealing with Odoo's verbose default responses.
Production Ready: Includes proper error handling, pagination for large datasets, and field validation that respects your Odoo model constraints.
Customer Service Workflow: Instead of logging into Odoo to check customer details during support calls, ask your AI assistant: "Find all recent orders for customer ABC Corp and their payment status." Get structured data instantly for better customer conversations.
Sales Pipeline Management: "Create a new opportunity for Microsoft worth $25,000 with expected close date next month, and assign it to Sarah from the Enterprise team." Your AI handles the record creation with proper field validation.
Inventory Monitoring: Set up AI-driven inventory alerts by asking: "Check which products have less than 10 units in stock and create a summary report." No custom dashboards needed.
The server runs locally on your development machine or as a containerized service. It connects to your existing Odoo instance via standard XML-RPC calls - no changes to your Odoo installation required beyond installing the companion module for access control.
For Claude Desktop users: Add to your MCP config and start querying immediately
For VS Code/Cursor users: Integrates with GitHub Copilot Chat for in-editor ERP access
For remote teams: Deploy via HTTP transport for shared access across your development team
Install via uvx (recommended):
{
"mcpServers": {
"odoo": {
"command": "uvx",
"args": ["mcp-server-odoo"],
"env": {
"ODOO_URL": "https://your-company.odoo.com",
"ODOO_API_KEY": "your-api-key",
"ODOO_DB": "your-database"
}
}
}
}
The server handles authentication, field validation, and permissions automatically. Your AI assistant gets structured access to business data without compromising security.
Security Note: All operations respect your existing Odoo user permissions and record rules. The API key determines what data is accessible, just like normal Odoo API usage.
This MCP server bridges the gap between conversational AI and structured business data, making your Odoo instance as accessible as asking a question. Perfect for developers who want their AI tools to understand and interact with their business systems directly.