Development + ProductionFull source availableSelf-Hosted

Develop and run Odoo with AI.
On your infrastructure.

Give developers, partners, and any AI agent a running Odoo environment โ€” with the code, database, tools, and lifecycle to take changes into production.

Your agent โ€” connected to Oduflow MCP

>

One operational layer

The same platform from first branch to production recovery.

Development and production have different safety requirements. Oduflow gives each its own lifecycle while keeping the workflow, infrastructure, and source code under your control.

01 / DEVELOPMENT

Give every branch its own Odoo.

branch: feature-invoice-flow

A complete Odoo workspace

database + filestore + repository + MCP

  • Isolated Odoo environments for every branch and developer
  • Production-sized database templates with copy-on-write filestores
  • Any agent, connected through MCP, CLI, or API to a running Odoo
Explore Development
02 / PRODUCTION

Run Odoo on infrastructure you control.

Push

Apply

Verify

History

  • A dedicated PostgreSQL cluster tuned separately from development
  • Verified deploys with automatic code rollback when health checks fail
  • Configurable S3 snapshots and cluster-wide point-in-time recovery
Explore Production
Dedicated production database cluster

Bring your intelligence

Any agent.
One Odoo workspace.

Choose the agent that works for you. Connect it through MCP, CLI, or API to the same running Odoo: code, database, filestore, logs, and tests. Your project stays ready when your choice of agent changes.

Claude Code, Codex, OpenCode, an agent in your editor, or your own automation. These are examples. The workspace is yours.

Connect your agent
Your agentYour editorYour automation

MCP / CLI / API

A running Odoo, ready for work.

  • Code and the correct branch
  • Database, filestore, and installed modules
  • Tests, logs, shell, and browser feedback
  • A context that belongs to the project

From a business need to production

One project. A complete delivery cycle.

The environment holds the context as the work moves between people, agents, and development and production operations.

  1. 01

    Start with real context

    Create a development environment from a template or an allowed production copy. Apply neutralization and your data sanitization rules.

  2. 02

    Build together

    Work on a branch with your agent. Invite a trusted client or partner into the running environment to reproduce and review the change.

  3. 03

    Test and review

    Apply modules, run tests, inspect logs, and check the Odoo UI. Your team decides when the change is ready to ship.

  4. 04

    Deploy and verify

    Deploy the code manually or through a GitHub webhook. Oduflow applies the update and checks the running production service.

If deployment verification fails: Oduflow attempts to restore the previous code and restart the service. Code rollback does not undo database changes.

If data needs recovery: restore a configured snapshot explicitly, or use cluster-wide point-in-time recovery. Routine code deployment preserves your live business data.

Explore deployment and recovery

Built for the way you work

Whoever you are in the Odoo ecosystem โ€” Oduflow removes the environment friction between you and shipped features.

Internal Odoo Teams

โ€œOur Odoo is business-critical, but development and production still depend on infrastructure only one person understands.โ€

  • Keep development data, source code, and production operations on company-controlled infrastructure
  • Give internal developers an isolated environment for every branch
  • Standardize deploy, backup, health, and recovery workflows in one platform
Explore production

Integrators & Service Teams

โ€œA dozen client databases, several Odoo versions, one overloaded staging server โ€” and disk space always running out.โ€

  • A named template and repeatable workflow per client, project, or Odoo version
  • Copy-on-write storage for parallel production-shaped environments
  • Separate teams and clients while sharing infrastructure efficiently
Explore development

AI-Enabled Engineering Teams

โ€œCoding agents write Odoo modules but can't install, test, or debug them โ€” a human babysits every iteration.โ€

  • MCP tools for environments, modules, tests, logs, shell, SQL, and production operations
  • A direct feedback loop: code, apply, test, inspect errors, and iterate
  • The same auditable infrastructure path for people, CI jobs, and agents
Browse the MCP toolkit

Visual dashboard included

A lightweight web UI for visibility and manual operations โ€” check status, browse logs, or trigger actions when you need hands-on control.

Why Oduflow?

Claude is the worker.
Oduflow is the workplace.

AI agents can already write Odoo code. A shared, repeatable environment lets clients, developers, and partners turn that capability into a delivery process that continues beyond one session.

Read why Oduflow exists

Up and running in minutes

Install Oduflow, connect any agent, and start working with a running Odoo. You can also work directly through the dashboard or CLI.

1

Install Oduflow

One package, batteries included. On first launch Oduflow sets up the Docker network and PostgreSQL automatically.

$ uv tool install oduflow
2

Connect any agent

Use MCP, CLI, or API from your preferred agent. This Claude Code command is one example; your workspace stays independent of the client.

$ claude mcp add oduflow -- oduflow
3

Ask for an environment

The agent handles containers, database cloning, module installs, and reads its own feedback from logs and tests.

> Create an Odoo 19 environment for this branch and run our module tests

Prefer a persistent server with a web dashboard and REST API? See the Quick Start

Technical discovery

Find the right plan for your setup.

Tell us how your team develops and operates Odoo today. We will map Oduflow to your infrastructure, security requirements, and rollout goals before proposing commercial terms.

Architecture first. We start with databases, repositories, deployment flow, and recovery requirements.

Your infrastructure. Source code, Odoo data, and operational control stay with your organization.

Terms that fit. Licensing, onboarding, support, and training are scoped in the written proposal.

Your Odoo landscape

Enough context for a useful first conversation

Frequently asked questions