QA

Create test cases and run QA tests before release.

The QA section helps you verify that Sagent responds safely and behaves as expected. It turns key shopper journeys and safety requirements into repeatable checks before release.

Use Test Cases to define expected behavior. Use Test Results to run those checks and investigate failures.

Use Test Results to run safety and regression tests. Use Test Cases to define the prompts and assertions those tests evaluate.

Use QA as a release gate

Test Cases define what must pass. Each case contains a shopper scenario, prompt, expected outcome, and any required assertions.

Test Results show what happened when Sagent ran those cases. Use them to compare actual behavior with the expected outcome and identify failures.

Run QA in this order:

1. Create or update Test Cases for new journeys, tools, content, and safety risks.

2. Run Test Results against the release candidate.

3. Investigate unexpected outcomes and correct the related configuration.

4. Run affected cases again and record the final result.

Publish only when required cases pass and unexpected results are resolved. Treat a changed catalog, prompt, knowledge source, translation, or design as a reason to rerun affected cases.

Passing an earlier run does not approve later changes. Run Test Results again after each release-impacting update.