API Test Generation
API Test Generation separates source selection, discovery, and test design. Generated scenarios are reviewed in a session and saved individually to AutoRunner.
1. Choose the source
| Tab | Use it for |
|---|---|
| Use Inventory | Select a published inventory version and the sources included in it |
| Upload File | Upload new API files for automatic analysis |
| Paste cURL | Add one or more individual requests with Add cURL |
Uploads support OpenAPI, Postman, HAR, WSDL, GraphQL, Protobuf, and AsyncAPI, up to five files of 10 MB each. If there is no published inventory, upload and analyze a source first.
Confirm the selected version and sources, then select Continue to Discovery.
2. Review discovery and scope
Discovery summarizes Endpoints, Services, Relations, and Sources. Inspect the Endpoints, Service Map, Flow, Relations, and Security Schemes tabs against your API contract. Open Inventory opens the inventory version behind this scope.
Services are selected by default. Expand a service to change its endpoint selection, or use the search and filters to find a smaller scope. At least one endpoint must be selected before Continue to Test Design is available.
3. Design the tests
| Method | What it does |
|---|---|
| AI Prompt Design | Describe the tests to generate in natural language |
| Import from Test Repository | Reuse existing test cases from your repository |
| Templates | Start from Sanity, Smoke, or Regression coverage |
For AI prompts, describe critical paths, negative cases, authorization behavior, and exclusions. Set Number of scenarios to the scope you can review. Keep credentials and private customer data out of the prompt.
Templates show their planned scenario count: Sanity Suite checks essential behavior, Smoke Suite covers critical paths, and Regression Suite adds negative and boundary cases.
Check Execute API requests during generation before starting. The workspace default is configured in Settings, and enabled sessions can send requests that change API data.
Select Generate Tests once the scope and design are ready.
4. Follow the session and review results
The session reports progress while scenario workflows and their automation are generated. Scenarios appear as they are produced. You can return through Sessions.
Open a candidate to review its ordered requests, assertions, captures, scripts, and variables. Verify paths and methods against the selected source version and correct the automation before saving. Generated Tests and Flows explains the review process.
Save scenario copies the reviewed candidate into an AutoRunner folder. Each save creates an independent scenario; the candidate remains available for further edits. Saved copies appear under Scenarios. Saving does not start an AutoRunner run.
