First 10 Minutes

This page gives you one small successful loop: launch, look around, make a safe edit, and checkpoint the state.

This is a quick win path

You can safely skip advanced deployment, Kubernetes, and storage configuration for now.

1. Open The Repository

Start in the repository folder:

pwd
ls

You should see files such as Dockerfile, docker-compose.yml, README.md, docs/, docker/, and scripts/.

2. Prepare Local Settings

Copy the example environment file if .env does not exist:

cp .env.example .env

Edit .env only for local settings and tokens. Do not commit .env.

3. Launch The Container

docker compose up --build

If you already built the image, this may be enough:

docker compose up

4. Open The Workspace Interfaces

Typical local links are:

Interface	What It Is For
OpenClaw Control UI	Chat, sessions, agents, and model routes
JupyterLab	File browsing, notebooks, markdown, figures, and previews
Workspace CMS	Reviewing private drafts and promoting public pages

Ports can vary when multiple instances are running. Use:

docker compose ps

5. Check Health

Run:

make doctor
scripts/status.sh
scripts/test-working-group.sh

These commands inspect the setup without asking agents to perform open-ended work.

6. Make One Safe Edit

Open PROJECT_CHARTER.md in the workspace and update the project title or purpose.

This is safe because it changes project memory, not credentials or container internals.

7. Checkpoint

At the end of a session, record what changed in the working group notes or checkpoint files. If the change belongs in the reusable template, commit it to git. If it is private project work, keep it in the mounted workspace or external storage.

make checkpoint

Do not commit secrets

.env, API keys, Slack tokens, private data, and credentials should stay local or in a secret manager.