All products

Prepare Postman test data with AI

Decide whether a single input file will do, or combine and filter several sources into a view the collection can run one row at a time.

Prepare Postman test data with AI when a collection needs the active products from one file and their prices from another. The datasets skill helps an agent decide whether to use a simple iteration file or create a queryable dataset. For a one-off run over an unchanged CSV, a saved view may add work without improving the result.

A join deserves a look before requests start. If a product has two matching prices, one input item can become two test runs. Inspecting columns, sample rows and the total makes surprises visible while the check is still cheap.

What it does

  • Feed one view row into each collection iteration.
  • Give a script access to a dataset without accidentally promising row-driven runs.
  • Compare the rows before and after joining two sources.
  • Check whether a teammate will receive a copied file or a machine-specific path.
  • Distinguish registering a local source from uploading its contents to a workspace.

Install and use

  1. Provide the official Postman CLI, an authorised collection and fictional input files. A live source also needs the correct connector and permission to read it.
  2. Ask, “Join these two sample files by product code, show the first five rows, then tell me how many iterations the collection should make.”
  3. Compare the completed run with the saved view and identify missing test cases separately.

FAQ

Source

Made by Postman. Original skill: datasets, published on GitHub under the Apache-2.0 licence.

View on GitHub