Skip to main content
Publish a note, a dashboard or a data frame from Desktop to the server, see what is published, and cancel the publication of a note or dashboard.

Before you start

  • Exploratory Desktop is running, the project is open (Choose a project) and you are signed in (expl status shows the signed-in user).
  • Publishing and cancelling work only when Allow create, edit and delete is on in System Settings > MCP / CLI.
  • Publishing makes the content readable by anyone who has the URL, and cancelling a publication cannot be undone. So both ask for confirmation first. In a terminal, expl asks Continue? [y/N]. From a script or an AI agent, add --yes, after you have checked the target. See Confirmation and long-running operations.

Steps

1. Find the target and the team

note list returns the note ids, and dashboard list the dashboard ids. team list returns the teams you belong to; a team name is only needed to publish into a team the first time.

2. Publish

--tags adds tags (repeat the flag for several), --team <team name> sets the team on the first publication, and --change-description writes the comment of a republication. Publishing again updates the same URL.
To publish a data frame, use expl dataframe publish <dataframe_id>. --step-name chooses the step to publish; without it, the last enabled step is used. expl published list shows the public URL (publishedUrl). If the result has outputOlderThanSource set to true, the published content is older than the current body. A dashboard can be published only after Desktop has rendered it. When No rendered output appears, open the dashboard once in Exploratory Desktop so that it renders, and run the command again.

3. See what is published

The result lists the notes, dashboards and data published from the project, with id, type, title, URL and team. --verify checks on the server that each one exists; it is slow, so use it only when you need it.

4. Cancel a publication

<id> is the id from expl published list. The public URL stops working. A published data frame is not cancelled by this command.

If it fails

For the arguments of any command, run expl <command> --help or expl tools describe "<command>". To do the same through MCP, see Publish in Use.