Odoo MCP Server for Claude

You have an MCP Server that exposes Odoo sales, customers, and CRM data to Claude AI, enabling real-time business analysis and more informed decisions. It connects to Odoo via API keys, runs analysis tools, and serves structured data through a simple HTTP interface that Claude can query to generate insights, dashboards, and proactive recommendations.

How to use

You connect your MCP client to the HTTP endpoint exposed by the server and request the available analysis tools. Use natural language prompts or structured requests to retrieve sales data, customer insights, CRM opportunities, product performance, and territorial analyses. You can combine data views across tools, such as pulling a territorial overview and then drilling into VIP customers in a specific province.

How to install

Prerequisites you need before installation are Docker and Docker Compose, an Odoo instance (v14+), and an Odoo API key to enable secure access.

# Copy the example environment file
cp .env.example .env

# Edit with your credentials
nano .env

ODOO_URL=http://tu-instancia.odoo.com/
ODOO_DB=tu-database-name
ODOO_USERNAME=integracion@tuempresa.com
ODOO_API_KEY=tu-api-key-de-odoo

Steps to deploy locally

Run the containerized server locally to get started. Build the image, start the services in the background, then monitor status and logs.

# Build the Docker image
docker-compose build

# Start the services in detached mode
docker-compose up -d

# Check the running containers
docker-compose ps

# Tail the logs for troubleshooting
docker-compose logs -f

Health check and endpoints

After the server is running, verify its health and available tools through simple HTTP checks. You can confirm the health, view available tools, and explore the API endpoints.

# Health check
curl http://localhost:8000/health

# List available tools
curl http://localhost:8000/tools

# Interactive API docs (Swagger)
open http://localhost:8000/docs

Claude integration notes

You can configure Claude to use this MCP server by adding a server entry to your Claude Code configuration. The following example shows how to point Claude to the local MCP server and use HTTP transport.

{
  "mcpServers": {
    "odoo": {
      "url": "http://localhost:8000",
      "transport": "http"
    }
  }
}

Available tools

get_sales_data

Fetches sales data with advanced filters for time ranges, customers, products, and regions.

get_customer_insights

Provides RFM-based segmentation and customer insights such as VIP, At Risk, New, Inactive, and Regular.

get_crm_opportunities

Analyzes the CRM pipeline and shows opportunities by stage, value, and probability.

get_product_performance

Ranks products by sales, revenue, and region to identify top performers.

get_sales_team_performance

Returns metrics for the sales team, including deals closed and revenue by rep and territory.

search_customers

Searches customers by name or criteria and returns matching results.

get_territorial_analysis

Provides territorial analyses by province and city, including revenue, orders, and customer coverage.

get_comprehensive_data

Offers a complete data analysis in a single call, aggregating multiple dimensions.