Before you begin#
- An account, and permission to create whatever docs calls a project or a workspace.
- Somewhere safe for a credential: a password manager, or your platform's secret store.
- Fifteen uninterrupted minutes.
Fill this in: list the real prerequisites — a supported runtime version, a minimum plan, an admin role. Readers forgive a long list at the top; they do not forgive finding out about one halfway through.
1. Create a project#
Sign in and create your first project. Give it the name your team already uses for this work rather than a test name — first projects have a habit of becoming production.
Checkpoint: the project appears in the sidebar under a name you recognise.
2. Connect it to something real#
A product like this only becomes useful once it is pointed at your own data. Do that now, with a small, low-risk source rather than your largest one.
export API_TOKEN="paste-your-token-here"
curl -X POST "https://api.example.com/v1/projects" \
-H "Authorization: Bearer $API_TOKEN" \
-H "Content-Type: application/json" \
-d '{"name": "first-project", "source": "sandbox"}'Fill this in: replace the host, the path and the fields above with a request that actually works, and say what a successful response looks like.
Checkpoint: the response carries an id, and the project shows the connection as active.
3. Do the thing the product is for#
Run the core workflow once, end to end, against that small source. Resist configuring anything else until you have seen one result.
Checkpoint: you can point at an output and say what produced it.
When a step fails#
Read the error text before changing anything — most setup errors name the field they object to. Then check the three usual suspects: the credential belongs to a different environment, the account is missing a permission, or a required field was left empty.
Next steps#
- What docs can do — what else is available now that it runs.
- Invite your team — add the people who will use it daily.
- FAQ — the questions that come up right after setup.