The biggest thing I learned from community operations is expectation management, so the most frequent instruction I issued to ChatGPT recently is: "Please help me optimize my expression from a more neutral perspective." I tested several mines today. From the test experience, I still like StandX very much. The interface display is very clear throughout the process, which slightly makes up for the pain of being stuck by bugs in another mine. But the tutorial I wrote is still very conservative. I look forward to the birth of the next good mine...