Skip to content

Step 3: Enable AI on a project

At this point you have a configured AI server and a prompt set. What remains is to associate them with a project so that AI-driven generation becomes accessible to the testers of that project.

Prerequisites

  • Step 1 completed: an AI server is declared and tested.
  • Step 2 completed: a prompt set is available.
  • SquashTM Ultimate 💎 license.
  • Administrator or project leader rights.

Access the project's AI configuration

Go to Administration > Projects, open the project's view page, then click the Artificial intelligence anchor AI icon.

AI configuration at project level

Fill in the three fields

The project's AI configuration relies on three fields:

  • AI prompt language: the language in which the prompt set is written. This value also determines the language of the requirement data that SquashTM injects into the prompts (for example the nature or the criticality of a requirement via {{requirement.nature}} and {{requirement.criticality}}).
  • Artificial intelligence server: the server declared at step 1.
  • Prompt set: the set chosen at step 2, whether it is a default system set or a custom set.

Language consistency

The selected language must match the language in which the prompt set was written, so that the instructions and the injected data remain consistent with each other.

Check tester permissions

For a user to be able to use generation on this project, they must have read permissions on the project's requirements and test case creation permissions on at least one project. If needed, check permissions from the project's Clearance anchor.

What you have accomplished

  • An AI server is linked to the project;
  • a prompt set is linked to the project;
  • the prompt language is defined;
  • authorized testers will see the Generate test cases action in the requirement menu of the project.

Next step

Generate test cases. This is the last step: it validates that the whole chain works.