The route

Generate the project docs

Every pipeline you've built in this course is already documented. You didn't write a word of it: Odara reads the project — pipelines, their diagrams, every node's configuration, the saved connections — and renders it as documentation that is never out of date, because it is the project. This level opens it, reads it, and adds the one thing Odara can't work out on its own: why a pipeline exists.

Goal

Find your course pipelines in the Documentation workspace, read one end to end, and give it a description that will travel into the PDF and the published site.

You'll need

  • The odara-academy project with the pipelines from the earlier worlds.
  • About 15 minutes.

Steps

1. Open Documentation

Editor menu ▸ Documentation. There's nothing to set up: the header already says how many items it documented.

Project Overview: counts, totals and the most recent pipelines

Project Overview gives the counts (pipelines, maestros, resources), the total number of nodes, and the most recently edited pipelines. The sidebar lists every pipeline, maestro and resource; the search box above it filters them all.

The page rescans when you come back to the tab or press Refresh, so it follows your edits in the Editor.

2. Read one pipeline

Click w1-4-filter-to-csv. Its page has:

  • metadata — nodes, edges, created, updated;
  • a Description box;
  • the Pipeline Diagram — the canvas itself, rendered automatically;
  • the nodes grouped by role — sources, transforms, targets — each with its configuration.

A pipeline's page, with its description and diagram

Expand the active_pro card: the SQL query is right there, as it runs. That's the level of detail a reviewer or a new teammate needs, without opening the Editor.

Node by node: the SQL that actually runs

3. Write the description

The diagram shows what; it can't show why. Type in the Description box and press Ctrl+S (Cmd+S on a Mac):

Keeps the active customers on the pro plan and writes them to active_pro_customers.csv for the sales team. Source: customers.csv (World 1). Rule: plan = pro AND active.

A good description answers three questions: where the data comes from, what rule is applied, and who uses the result. It's stored on the pipeline, so it shows up in the Editor, the PDF and the published site.

4. Look at a database pipeline

Open w4-2-read. The connection is documented — host, database, user — but the password is replaced by ***, here, in the PDF and on published pages.

Expected result

  • Documentation lists all your course pipelines and your saved connection.
  • w1-4-filter-to-csv shows its diagram, its three nodes and your description.
  • Database nodes show postgres://user:***@host… — never the password.

Common pitfalls

A pipeline shows no diagram. Diagrams are captured from the canvas the first time the page is opened. Wait a moment, or hover the diagram and regenerate it after changing the pipeline.

The description disappears. It wasn't saved — the status line says Unsaved changes until you press Ctrl+S or Save.

Surprised by what's in there. Everything in the project is documented, test pipelines included. Delete what shouldn't be read by anyone else — or choose what to publish in level 8.3.

What you built

Documentation you didn't have to write and don't have to keep up to date. The next two levels get it out of Odara: as a PDF, and as a link.

Check yourself

One question, drawn from a larger bank. Get it right to pass.

  1. What does a database node's documentation show for the password?