Skip to content

AI Test Case Generator Overview

2 min read

The AI Test Case Generator drafts test cases from what you already have: requirements, notes you paste in, or a specification document.

It runs in two stages. First it maps coverage — a broad list of one-line scenarios you can scan. Then it writes full test cases, with steps and expected results, only for the scenarios you tick.

Nothing is saved until you add it.


A finished test case costs roughly ten times what a one-line scenario costs to produce. Within the same budget, a single request can write about ten complete test cases, or map about forty scenarios.

Forty scenario titles is something you can read in a minute and check for holes. Ten finished test cases is not — to spot what they miss, you would have to already know what should have been there.

So the map comes first, and the map is the thing you approve.

The AI coverage map: sources on the left, and on the right the scenario counts, requirement coverage, gaps, per-requirement depth and the scenario list grouped by focus.
Stage one: the coverage map, with counts, per-requirement depth, and the requirements nothing was proposed for.

Both entry points use the Draft with AI button.

  • Repository — opens the full coverage workspace, where you can combine several requirements, pasted context and documents in one brief. This is the flow the rest of this section describes.
  • A requirement’s detail page — opens a smaller, single-step version that drafts test cases for that requirement directly, without the coverage map. See Draft from a requirement.

A brief can combine any of:

  • Requirements you select from the project.
  • Additional context you paste in — rules or edge cases the requirements don’t state.
  • Documents you attach — PDF, Word, Markdown or plain text.

Everything you supply is used as source material for one run. Nothing is uploaded to storage, and the generator never reads your repository, your issue tracker, or anything you did not attach.


Accepted test cases are ordinary test cases. They are active from the moment they land, get their own test case code, sit in the folder you chose, and link to the requirements they cover.

They carry one AI tag so you can always tell where they came from, and a Generated by AI panel on the test case showing the scenario and lens that produced them. There is no draft, review or approval state to move them out of.


  • It does not generate requirements, test suites, releases or defects — test cases only.
  • It does not read your repository, your Jira, or any source you did not attach.
  • It does not save anything on its own, run in the background, or regenerate on a schedule.
  • It does not invent detail. Where the source is too thin it returns fewer scenarios and says so, rather than padding the list.