OOMOL FLOW

Let your AI agent build the workflow.

Create, check, and run a Flow through oo flow, then inspect and edit the same graph in Workbench. Use OOMOL's official hosted version or deploy Open Flow on your own infrastructure.

Use OOMOL Flow Self-Host Open Flow

Read the Guide

Open Flow Workbench showing a Gmail-to-Feishu workflow, connected nodes, and successful run results. — Gmail to Feishu: connected nodes and execution results in the Open Flow Workbench. Watch the Demo

Build it together. Inspect every step.

One Flow for you and your agent

Agents use oo flow to create and check a Draft. Open the same Flow in Workbench to inspect its graph and continue editing.

Visual structure, custom JavaScript

Connect typed inputs and outputs on the canvas. Use Code Tasks when a step needs custom logic.

Run, inspect, and publish

Check the Draft, inspect node outputs and run history, then publish when ready. Start manually or with Cron, Webhooks, polling, and Provider events.

Choose where your Flow runs

The hosted and self-hosted paths use the same Flow product and Workbench. Choose who operates the deployment and its connected services.

Official hosted version

OOMOL Flow

Open Workflows in OOMOL Console. OOMOL operates the deployment and provides managed OAuth apps for supported integrations.

Use OOMOL Flow

Self-hosted · Beta

Open Flow

Run the open-source server with Docker or deploy the image to Fly.io. You manage storage, backups, upgrades, tokens, and Connector configuration.

Self-Host Open Flow

Open Flow is in beta and has not reached its first stable release. Check the repository's current deployment requirements before installing.

Choose where your Flow runs
Responsibility	OOMOL Hosted	Self-Hosted Open Flow
Deployment and upgrades	Operated by OOMOL	Managed by your team
Persistent run data	Stored in the hosted deployment	Stored in the storage you configure
App connections and OAuth	Managed OAuth apps for supported integrations	Choose a Connector and configure the required accounts and OAuth apps
Authoring and inspection	oo flow and Workbench	oo flow and Workbench against your deployment

How Flow, Skills, and Connector fit together

Flow runs the automation

Nodes and edges define dependencies; the deployment owns runs, publications, and triggers.

Skills describe reusable methods

Skills give an agent task instructions. Flow provides a persisted graph and execution lifecycle for automation.

Connector calls the Apps

Third-party Actions and Provider Triggers use a Connector runtime. Provider credentials stay with the Connector; Flow stores connection identities.

Start with a task you already do

Build a workflow that reads unread Gmail messages, formats them, and sends them to Feishu.

Connect the required accounts, let the agent create a Draft, inspect the result, and publish only when you are ready.

Use OOMOL Flow Read the Guide