Step-quality suggestions, executable Test Plans, and Thunders AI chat feedback
TL;DR
Pre-run and post-run step-quality suggestions are now generally available, with a new Thunders MCP tool to apply them in bulk across test cases. This release also lets Test Plans run just like Test Sets with their own owner, adds thumbs up/down feedback on Thunders AI chat messages, progresses the Mobile Testing private beta, and ships a test-case-level accessibility report alongside reliability fixes.
FEATURE
- Pre-run and post-run suggestions are generally available: get step-quality suggestions and ready-to-apply rewrites before and after a run, for more stable tests.
- Thunders MCP: apply pre-run and post-run suggestions in bulk: read step-quality suggestions across many test cases at once, apply the ones that come with a ready rewrite, and leave the rest for a person to decide.
- Test Plans can now be executed: Test Plans now inherit the same structure as Test Sets: they can be executed, they get an owner, and they show their own test cases.
- Test Set owners: every test set now has an owner, shown in the list and its properties panel, just like test cases. The creator becomes the owner by default and you can change it.
- Thumbs up/down feedback on Thunders AI: rate a Thunders AI message right from the chat.
- Mobile Testing (private beta): continued progress on the mobile testing beta.
- Device Pools Management: create Android or iOS device pools from the Devices page, and select a specific pool at execution time.
- New mobile execution settings: set connectivity and device permissions for a run.
- Test case level accessibility report: the accessibility persona can now produce one accessibility report for the whole test case, instead of a separate report per step, when its reporting mode is set to comprehensive.
IMPROVEMENT
- Apps: screenshots now live at the app level. The old project-level screenshot code has been removed.
- Model availability signal: see when the AI model behind a step is temporarily unavailable instead of a run just failing.
- Test Run Console Logs: now disabled when a run has no console output, instead of downloading an empty file.
FIX
- Drag-and-drop reliability: clicks and drags now land on the correct target, including native HTML5 drag widgets and vision-based coordinate fallback.
- Duplicate step execution: a discovery step no longer runs twice by replaying a stale result from a previous attempt.
- Step-edit conflicts: we only flag a conflict when someone else actually changed the test case, not on your own concurrent edits.
- Screenshot capture failures: a failed screenshot capture now shows a clear error instead of a generic "model unavailable" message.
- Natural language step recognition: random-value generation, exact-word-count phrases, wait-for-disappear, and off-enum AI issue types are now classified and handled more reliably.
- Visual Compare and accessibility scans: improved detection for color-blind-safe comparisons and high-DPI baselines, clearer guidance when a baseline needs rescaling, and accessibility scans now run at the right time to match the screenshot.
- Downloads and mobile API steps: CDP downloads now honor your HTTPS-error setting, and mobile API-call steps return the real HTTP status instead of a placeholder.
- UI polish: fixed devices tab and test-run empty states, the Thunders AI page URL, and a copy issue when reading an uploaded file in Thunders AI.
Ready to ship faster with smarter testing?
