Odoo 17.0 Community and Enterprise

Governed AI access to your Odoo data.

Foggy Odoo Bridge Pro exposes approved Odoo business data to MCP clients and the built-in Odoo AI Chat through a governed semantic query layer. It is built for analysis, reporting, and operational questions while preserving Odoo access rules, multi-company boundaries, and field governance.

Read Setup Guide Book a Demo Review Privacy
Embedded engine by default MCP endpoint for AI tools Optional built-in AI Chat

What It Does

The addon gives approved AI clients a controlled way to ask questions over Odoo data. Queries are executed through semantic models, not unrestricted database access.

MCP access

Connect tools such as Claude Desktop, Cursor, VS Code extensions, or other MCP clients to /foggy-mcp/rpc with an Odoo-issued API key.

Built-in query engine

The recommended setup runs the governed query engine inside Odoo. No separate Foggy server is required for day-one embedded deployment.

Business-oriented metadata

AI tools receive model and field descriptions for Odoo Sales, Purchase, Accounting, Inventory, CRM, HR, Contacts, Companies, and Products.

Security And Governance

Foggy Odoo Bridge Pro keeps Odoo as the authority boundary. The AI client only sees the data surface allowed for the authenticated Odoo user.

Odoo-native permissions

  • ir.model.access controls which Odoo models a user can query.
  • ir.rule domains are converted into server-side SQL conditions.
  • Multi-company access is enforced before query results are returned.

Field-level governance

  • Admins can block or mask sensitive fields per model.
  • Blocked fields are removed from metadata, queries, AI Chat, and exports.
  • Masked values stay masked in query results and exported files.

SQL evidence

Query responses can include SQL evidence for debugging and review, helping administrators understand what was executed without exposing unrestricted SQL execution to users.

Audit-aware workflow

AI Chat tool calls, exports, and governed query activity are designed to fit operational review and support workflows.

Optional Built-in AI Chat

Teams that want an in-Odoo experience can enable AI Chat. Teams that only need MCP access can keep AI Chat disabled.

Provider choice

Supports OpenAI-compatible endpoints and Anthropic-format providers when the optional provider SDKs are installed and configured.

Explicit opt-in

External LLM data transfer is disabled until an administrator enables the AI Chat consent setting and configures the provider.

Visible query process

Admins can inspect tool calls, query payloads, SQL evidence, and exports for review and troubleshooting.

AI-generated analysis is not a substitute for business review

AI answers may be incomplete, incorrect, or may not fully match your intended business definitions. Important decisions should be checked against source records, tool calls, query payloads, SQL evidence, and exported results.

Supported Odoo Surface

The public release baseline includes 12 core query models. Project and Manufacturing models are available when the corresponding Odoo apps are installed.

12core query models
2optional app models
4MCP tools

Core models

Sales Orders, Sales Order Lines, Purchase Orders, Customer Invoices/Bills, Journal Items, Payments, Inventory Transfers, Employees, Contacts, Companies, CRM Leads, and Products.

Optional models

Project Tasks and Manufacturing Orders are exposed only when the related Odoo applications are installed in the target database.

Setup Path

The default path is intentionally short: install the addon, keep Embedded mode, create user API keys, and connect MCP clients to the Odoo endpoint.

1. Install

Install the addon in Odoo 17.0 and ensure required Python dependencies such as asyncpg and openpyxl are available.

2. Configure

Open Settings -> Foggy MCP, verify Embedded mode, run the health check, and configure optional AI Chat provider settings only if needed.

3. Connect

Create API keys for authorized users and configure MCP clients with https://your-odoo.com/foggy-mcp/rpc.

Need the exact steps?

The setup guide covers dependencies, Embedded mode, API keys, MCP clients, AI Chat opt-in, governance, and troubleshooting.

Open Setup Guide

Current Boundaries

These boundaries are stated plainly so customers can evaluate the release accurately.

Supported now

  • Odoo 17.0 Community and Enterprise.
  • PostgreSQL-backed Odoo databases.
  • Embedded query engine as the recommended deployment mode.
  • MCP clients and optional in-Odoo AI Chat.

Not guaranteed

  • All complex analyses being correct on every LLM model.
  • MySQL as an officially validated target for this Odoo bridge release.
  • External MCP clients receiving the same internal AI Chat system prompt.
  • Skill/fsscript high-level analysis playbooks in this release.