Joule test case generation in SAP Cloud ALM

SAP has built test generation into SAP Cloud ALM itself. Its Help Portal documents Generating Manual Test Cases with Joule, which uses two Joule agents to turn process documentation, requirements and user stories into a test plan with manual test cases. Before planning around it, read the status: SAP marks the feature as part of an SAP Early Adopter Care (EAC) program, and features in an EAC program are not available to customers outside it. Check the current Help page before you plan a project around it; the status will change.

Two stages, two confirmations

The workflow runs in one Joule conversation and stops twice for you. In stage one, the Test Scoping Agent analyzes a process and its requirements and user stories and proposes a Scope Summary: relevant requirements and user stories, process paths (main flows) and process variants (alternative or exception flows), plus related information kept as reference. Nothing continues until you confirm it. In stage two, the Test Preparation Agent proposes a Test Plan Summary with planned test cases and coverage categories such as happy path, error case and edge case. After you confirm that, it generates the test cases asynchronously and notifies you in Cloud ALM, and by e-mail if configured.

What it needs

SAP also lists usage considerations that shape how you plan. Each run works on one Solution Process Flow Diagram. The Test Scoping Agent follows direct links only; requirements linked indirectly are ignored. Only manual test cases are generated; automated test scripts are not, in this release. The agent can't see what is on your screen, so name the process, requirement or user story in the prompt. And output quality reflects the quality and completeness of your requirements, user stories and process documentation.

What you get, and how it is marked

The result in Test Management is a new test plan, manual test cases with detailed steps and expected results covering the confirmed paths and variants, traceability links back to the requirements, user stories and business processes, and coverage information. Each generated test case is marked Manually created with AI on its General Information tab and shows an information banner while it is In Preparation, saying it was generated with AI and should be reviewed before use. The banner is a reminder, not a review: the case still needs the same review as one a consultant wrote, before anyone sets it to Prepared.

The neighbors: requirements and test data agents

Two related Joule agents in Cloud ALM matter to testers. The ALM Requirement and User Story Agent helps generate requirements and user stories, for example from workshop transcripts, and find and update them. Better requirements make better generated tests, and the same review rules apply. The Test Data Management Agent, also an EAC feature, analyzes the instruction text of manual test cases and returns the SAP S/4HANA business objects and field parameters each one needs, grouped by category, as a basis for sourcing test data. Lesson 4 picks up that idea.

What would this mean for Contoso? Its S/4HANA program runs in SAP Cloud ALM, but its process designs mostly reuse SAP Best Practices content rather than custom solution processes, so the prerequisites would force a decision: build custom flow diagrams for the processes it wants generated, with requirements linked directly. That is a design-quality investment worth making for high-risk processes, such as procure-to-pay with the $25,000 purchase order approval, even before generation is available.