What is Model Context Protocol (MCP)?
Model Context Protocol (MCP) is an open standard created by Anthropic that defines how AI agents, LLMs, and automation tools connect to external systems. Think of it as a universal plug — one standardised way for any AI model to talk to any application.
Before MCP, every AI integration was custom-built. Want Claude to query your CRM? Build a custom API wrapper. Want ChatGPT to read your inventory data? Build another wrapper. Every integration was one-off, fragile, and hard to govern.
MCP changes that. It defines a standard protocol so AI agents can:
- Automatically discover what data sources and tools are available
- Read structured data from any connected system
- Perform actions (create, update, delete) in those systems
- All with governed, auditable, and scoped access
How MCP works — the technical basics
An MCP implementation consists of two parts:
MCP Server
The application that exposes data and actions via the MCP protocol. In our case, this is Odoo — the KSROlabs MCP Server module turns Odoo into an MCP-compliant endpoint.
MCP Client (AI Agent)
The AI model or agent that connects to the MCP Server. Claude, ChatGPT, LangChain agents, and any MCP-compatible tool can act as a client and query Odoo through the protocol.
When an AI agent connects to an MCP Server, it first calls initialize to discover the server capabilities. Then it can discover available tools (like list_models, search_records) and call them to interact with your Odoo data — all via a single POST /mcp endpoint.
Why use MCP for Odoo specifically?
Odoo is a powerful ERP — but connecting AI agents to it has historically been complex. Odoo has its own JSON-RPC API, complex access control lists, and many interrelated models. Getting an AI agent to query Odoo reliably, safely, and with proper governance required significant custom development work.
The KSROlabs MCP Server for Odoo solves this by implementing MCP natively inside Odoo as a standard module. You install it from the Odoo App Store, configure API keys with model-level access scopes, and any MCP-compatible AI agent can immediately start querying your Odoo data — with full Odoo permission enforcement and audit logging.
Governed access
Odoo RBAC is enforced for every AI action. Agents see only what their scoped API key permits.
Full audit trail
Every AI action is logged via Odoo chatter. Full visibility for compliance and security reviews.
4x throughput
Schema caching makes model discovery 96% faster. Production-ready at 200+ requests/sec.
What can you do with Odoo MCP Server?
Once your AI agent is connected to Odoo via MCP, a wide range of use cases become possible:
Getting started with Odoo MCP Server
Setting up the KSROlabs MCP Server for Odoo takes under 30 minutes:
Install from Odoo App Store
Search for "MCP Server" in your Odoo Apps menu or visit the Odoo App Store. Install like any standard Odoo module — no external dependencies.
Enable the MCP endpoint
Navigate to Settings → MCP Server. Enable the endpoint and assign the MCP Server Administrator role to trusted users.
Create scoped API keys
Create one API key per AI integration. Configure which Odoo models each key can access. Copy the token (shown once).
Connect your AI agent
Point your MCP-compatible client at your Odoo URL with the Bearer token. Claude, ChatGPT, LangChain — all connect the same way.
Free setup support included: KSROlabs provides free, expert setup guidance for every purchase. We help you configure scoped API keys, harden security, and connect your first AI agent — at no extra cost. Get in touch →
Frequently asked questions
What is Model Context Protocol (MCP)?
Model Context Protocol (MCP) is an open standard created by Anthropic that defines how AI agents, LLMs, and automation tools connect to external systems. It gives AI models a structured, governed interface to read data and take actions in real applications — like your Odoo ERP.
Why use MCP instead of a custom REST API for Odoo AI integration?
MCP is designed natively for AI agents — tools are self-describing, so agents automatically discover what they can do. A custom REST API requires manual documentation, wrapper code, and separate governance. MCP with the KSROlabs server gives you all of this out of the box, with Odoo permission enforcement built in.
Which AI agents work with Odoo MCP Server?
Any MCP-compatible AI client works with the KSROlabs Odoo MCP Server, including Claude (Anthropic), OpenAI ChatGPT with custom tools, LangChain agents, and any custom agent built on the MCP protocol.
Is the Odoo MCP Server free to use?
The MCP Server for Odoo is a paid module available on the Odoo App Store. KSROlabs includes free setup support with every purchase — so your first production deployment is fully assisted at no extra cost.
What Odoo versions does the MCP Server support?
The KSROlabs MCP Server supports Odoo 16.0, 17.0, and 18.0. Multiple version packages are available on the Odoo App Store.