odoo

This MCP server provides an agentic bridge to an Odoo ERP instance. It exposes a set of tools for CRUD operations, data introspection, and BI-style analytics that you can invoke from an AI assistant to read, write, and analyze Odoo data within the context of your business processes. Use the included tools to search and read records, create or update data, and examine model schemas, views, security rules, and workflows in real time. The server also supports advanced BI operations like read_group for aggregated reporting and export_to_csv for structured data export, enabling AI-driven reporting and KPI generation.

To use the server, configure your MCP client (IDE, editor, or integration) with the odoo-server MCP configuration and provide the required environment variables (ODOO_URL, ODOO_DB, ODOO_USERNAME, ODOO_PASSWORD). Once connected, you can invoke prompts or commands such as search_records, read_records, get_model_fields, read_group, and export_to_csv. The guided prompts available (analyze_sales, diagnose_inventory, audit_permissions, financial_overview) offer pre-trained workflows to perform structured analyses and present insights like revenue trends, stock health, and security posture directly through the AI client.

Prerequisites:

• Python 3.8+ installed on your system
• pip available for installing the odoo-mcp-server-conn package
• Access credentials to an Odoo instance (URL, database, user, and API key or password)

Installation steps:

1. Create a Python virtual environment (recommended): python -m venv venv source venv/bin/activate # macOS/Linux venv\Scripts\activate # Windows

2. Install the MCP server package (ODDO MCP Server Conn): pip install odoo-mcp-server-conn

3. Verify the CLI entry point is available (odoo-mcp-server-conn): odoo-mcp-server-conn --help

4. Prepare your environment variables (ODOO_URL, ODOO_DB, ODOO_USERNAME, ODOO_PASSWORD) for secure usage. You can export them in your shell or use a .env file with a tool that loads env vars before starting the server.

5. Run the MCP server locally for testing: python -m odoo_mcp_server_conn

6. Integrate with your MCP client by configuring the mcp_config as shown in this README. For production, consider using a process manager (e.g., systemd, pm2) to keep the server running.

Tips and caveats:

• Credentials are read exclusively from environment variables; avoid hardcoding passwords in configuration files.
• Use an API key instead of a password when available, to improve security.
• The server relies on Odoo's XML-RPC layer; permissions are enforced by Odoo and cannot be bypassed by the MCP server.
• If you encounter connection issues, verify ODOO_URL and that the target Odoo instance accepts external XML-RPC requests.
• In Claude Code, VS Code, Windsurf, and other integrations, ensure the env vars are correctly supplied in the respective MCP configuration blocks.
• The BI tools (read_group, export_to_csv) have limits (e.g., export_to_csv supports up to 1,000 rows); plan larger exports with iterative queries if needed.