Connect AI agents to Odoo — safely, with governance. Expose one stable endpoint (POST /mcp) with scoped API keys, permission enforcement, validation, and audit logs. The definitive MCP integration for Odoo.
The 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.
The KSROlabs MCP Server for Odoo implements this protocol inside Odoo, exposing your ERP as a secure, permission-aware endpoint for AI agents. No custom API development. No prompt injection risks. Just governed Odoo MCP integration that works out of the box.
Every feature is designed to keep your Odoo production instance safe while giving AI agents the access they need.
Single endpoint
One stable entry point (POST /mcp) for tool discovery and tool execution. Keeps integrations consistent and easy to maintain across AI clients.
Scoped API keys
Create keys per integration and restrict by model access. Rotate keys safely without breaking your Odoo permission model.
Audit-ready logs
Track which key called what, when, and the payload — helpful for security reviews, compliance, and debugging every AI action.
Field & record guards
Add per-key domain filters and optional field allowlists for precise, least-privilege access. AI agents see only what you allow.
Rate limits & IP rules
Global and per-key rate limits with optional IP allow/deny lists reduce abuse risk and protect your Odoo production instance.
Payload safety
Configurable payload size limits and optional error redaction keep integrations safe and stable in production environments.
Built for high-scale production deployments with comprehensive security governance, enterprise-grade performance, and horizontal scalability across unlimited workers.
Distributed Rate Limiting
Zero-Downtime Token Rotation
Schema Caching
Prometheus Monitoring
Enterprise Guarantee: Production-ready with 4x API throughput, 96% faster requests, zero-downtime token rotation, and multi-worker scalability. Tested and verified in real-world deployments.
Configure once, then integrate safely with any MCP-capable client. Get from installation to a working Odoo MCP integration in minutes.
Install the module
Install the KSROlabs MCP Server for Odoo from the Odoo App Store like any standard Odoo module — no external dependencies required.
Enable MCP Server in Settings
Navigate to Settings → MCP Server and enable the MCP endpoint. Grant the MCP Server Administrator group to trusted users.
Create a scoped API key
Create an API key in the MCP Server settings. Copy the token (shown once). Configure which Odoo models this key can access using the Key Model Access settings.
Set optional security rules
Optionally configure IP allow/deny rules, rate limits per key, field allowlists, and domain filters for precise least-privilege access.
Connect your AI agent
Point your MCP-compatible client (Claude, GPT, LangChain, etc.) at your Odoo MCP endpoint. Use the Bearer token for authentication.
Query and act on Odoo data
Your AI agent can now read, create, update, and delete Odoo records — all within the access boundaries you defined, with every action logged.
Eight purpose-built tools cover every AI agent use case — from model discovery to full CRUD operations. All tools are rate-limited and cached for performance.
list_modelsDiscover allowed models for a key (cached)
describe_modelFields and metadata for agent planning (cached)
search_recordsDomain-based search with validation
get_recordFetch by ID with field access control
create_recordCreate with validation and audit logging
update_recordUpdate with Odoo security rules enforced
delete_recordDelete only if allowed by key scope
initializeServer info, capabilities, and metrics
Batch support: JSON-RPC array payloads are supported for multi-call requests. All tools are rate-limited and schema results are cached for 5 minutes.
This release brings production-ready infrastructure improvements for enterprise deployments.
Why the Model Context Protocol is a better approach than custom REST APIs for connecting AI agents to Odoo.
| Feature | MCP for Odoo | Custom REST API |
|---|---|---|
| Protocol standard | Open standard (Anthropic MCP spec) | Varies per implementation |
| AI agent native | Yes — designed for LLMs | No — requires wrapper code |
| Tool discovery | Automatic — agents discover available tools | Manual — agents need documentation |
| Odoo permission inheritance | Built-in with KSROlabs | Must be implemented manually |
| Audit trail | Automatic via Odoo chatter | Must be added separately |
| Setup complexity | Install module, configure scopes | Build endpoints, auth, mapping |
We're not just a tool provider — we're your Odoo AI consultancy partner. We help organisations unlock the full potential of AI within their Odoo ecosystem, from MCP Server implementation to enterprise-scale AI orchestration.
MCP Strategy & Planning
Design your Odoo AI integration strategy. Identify high-impact AI use cases, scope implementation, and plan rollout across teams.
Implementation & Setup
End-to-end MCP Server deployment. Configuration, API key scoping, security hardening, and production readiness validation.
AI Agent Customization
Build and customize AI agents (Claude, ChatGPT) for your Odoo workflows — sales automation, CRM intelligence, finance processing, and more.
Security & Governance
Audit API key scopes, implement least-privilege access, enforce rate limiting, and establish compliance workflows for AI operations.
Ongoing Support & Optimization
Performance tuning, monitoring setup, token rotation management, and team training for sustained AI-driven productivity.
Installing the module is easy. Setting it up correctly for production AI usage is critical. We personally help you configure, secure, and deploy your MCP Server — at no extra cost.
What we help you with (FREE)
Why contact us?
MCP is powerful — but misconfiguration can limit its capabilities. When implemented correctly, this module can:
Install the MCP Server for Odoo or contact us for free setup support and expert guidance.
Common questions about the Odoo MCP Server, Model Context Protocol integration, and connecting AI agents to Odoo.
MCP is an open protocol created by Anthropic that standardises how AI agents and large language models (LLMs) connect to external data sources and tools. It provides a governed, structured interface so AI systems can securely read data and take actions in business applications like Odoo ERP.
Any MCP-compatible client works with the Odoo MCP Server, including Claude (Anthropic), OpenAI GPT models, and custom AI agents built on frameworks like LangChain, AutoGen, CrewAI, or the Claude Agent SDK. Any tool that speaks the MCP protocol can connect.
Yes. The KSROlabs MCP Server enforces Odoo's built-in access control lists (ACLs) and record rules. AI agents can only read or write data that their associated Odoo user account is authorised to access. Every action is logged in Odoo's audit trail.
Both read and write operations are supported. You configure per-client permission scopes — for example, allowing reporting-only (read) access for one AI agent and full CRUD access for another. Write access can be disabled entirely for maximum safety.
Install the module from the Odoo App Store like any standard Odoo module. After installation, navigate to Settings → MCP Server to configure client credentials, allowed models, and access scopes. No custom API development is required.
A REST API exposes raw endpoints that require custom integration code. MCP provides a standardised protocol specifically designed for AI agent interaction — with built-in tool discovery, structured responses, and governed access. MCP lets AI agents understand what actions are available without custom code.
Yes. Claude by Anthropic natively supports MCP. Once the KSROlabs MCP Server module is installed in Odoo, you can point Claude Desktop, Claude Code, or the Claude API at your Odoo MCP endpoint to query sales orders, invoices, inventory, CRM data, and more.
Yes. The MCP Server for Odoo works with both Odoo Community and Odoo Enterprise editions. All standard Odoo models are accessible, and the module respects the same permission system regardless of edition.
The MCP Server module is built and tested on Odoo 16.0 and above, including Odoo 17 and Odoo 18. Check the Odoo App Store listing for the latest version compatibility.
Yes. Each AI agent is configured as a separate MCP client with its own credentials and access scopes. You can run multiple agents simultaneously — for example, one for reporting and another for workflow automation — without any conflict.
AI agents can access any Odoo model you configure: sales orders, purchase orders, invoices, inventory, CRM leads, contacts, projects, timesheets, and custom models. You control exactly which models and fields each MCP client can see.
Yes. The module is designed for production environments with governed access, audit logging, configurable rate limits, and Odoo permission inheritance. It is production-safe by design and used by businesses connecting AI agents to live Odoo instances.
Beyond MCP — analytics connectors for Power BI, Tableau, Google Sheets, BigQuery, plus Google Drive storage and Google Shopping integrations for Odoo.