Runs

Execute generated tests in cloud sandboxes and read the results.

Run a feature's generated Playwright tests in an isolated cloud sandbox, then read the results digest and download artifacts.

Operation	What it does
run_feature	Dispatch a run.
run_suite	Dispatch runs for several features (or every generated one) at once.
wait_for_run	Block until it finishes (MCP convenience).
get_run	One run's status + results digest.
diagnose_run	A failed run's full failure context in one call.
list_runs	Recent runs, newest first (optionally by status).
get_run_artifacts	Short-lived download links for traces, screenshots, videos.

The dispatch → poll pattern is described in Long-running operations.