Let’s look at the conclusion first
Go to Feedback in Settings and fill in the required problem description; add up to 3 images and select recent AI diagnostic sessions to include. The application only generates the sharing package locally, and you decide whether to send it and to whom in the system sharing panel.
This route does not upload to the product-feedback reward task and does not grant membership. To join the three-day reward task, open My > Membership and Rewards > Earn Membership by Completing Tasks.
It is recommended to fill in like this
- Write clearly which page it takes place on, whether to use full screen or floating, constant speed or AI.
- Describe the operation by time, such as "switch to AI - open the system camera - read two sentences - show that the microphone is being restored."
- Write out what you expected to see, what actually happened, and whether the problem reoccurred every time.
- Add up to 3 screenshots that illustrate the phenomenon. The app re-encodes the image and removes the original metadata.
- The last 3 diagnostic audio sessions are listed by default; only the sessions related to the problem are kept, and those you do not want to provide can be deselected.
- After generating the sharing package, check the system sharing panel and proactively select email, instant messaging, or other delivery methods.
What a diagnostic session may include
Selected conversations may contain corresponding full transcripts, hotwords, recognized text and tags, algorithm events, and microphone audio. This information can help reproduce the issue, but may also contain private content, so please confirm your selection before sending. Unselected sessions will not be added to the sharing package.
Privacy and Cleanup
The app does not automatically upload feedback. The sharing package will be temporarily stored in the local sharing cache, and expired content will be cleared after about 24 hours; after the system sharing panel is opened, you still control whether it is actually sent.