# Run your first flow · Open Flow

Source: https://openflow.run/en/quickstart/

[Overview / Quick start](https://openflow.run/en/)

# Run your first flow

Start Open Flow with Docker, open the Workbench, and connect your AI agent to the same deployment.

## 01 · Start the server

Install Docker and OpenSSL, then build and run the server. The named volume keeps project data and run history across container restarts.

Shell

```
git clone https://github.com/oomol-lab/open-flow.git
cd open-flow

export OPEN_FLOW_TOKEN="$(openssl rand -hex 32)"
docker build --file apps/server/Dockerfile --tag open-flow-server:dev .
docker run --rm \
  --publish 3000:3000 \
  --env OPEN_FLOW_TOKEN="$OPEN_FLOW_TOKEN" \
  --volume open-flow-data:/data/open-flow \
  open-flow-server:dev
```

[Docker / GHCR ↗](https://github.com/oomol-lab/open-flow/blob/main/docs/server/docker-ghcr/README.md)

## 02 · Open the Workbench

Visit http://127.0.0.1:3000 and sign in with OPEN\_FLOW\_TOKEN. Create a project, add typed nodes, connect their inputs and outputs, then check and run the draft.

## 03 · Connect your agent

Install the oo CLI using its official guide. Set these variables in the shell running your agent so the CLI and Workbench use the same deployment.

[Install oo CLI ↗](https://github.com/oomol-lab/oo-cli)

Shell

```
export OO_OPEN_FLOW_URL=http://127.0.0.1:3000
export OO_OPEN_FLOW_TOKEN="$OPEN_FLOW_TOKEN"
```

[Open Flow + OpenConnector ↗](https://github.com/oomol-lab/open-flow/blob/main/docs/server/self-hosted-stack/README.md)

Before production use, configure TLS, backups, resource limits, and connector capabilities. External actions and LLM tasks require their corresponding host capabilities; they fail closed when these are absent.

Open Flow is in beta. Contracts are versioned; the first stable release is still ahead.