A software development team is using a large language model to sort user feedback for their new video editing application. They want to classify each piece of feedback into one of three custom categories: 'UI/UX Suggestion', 'Bug Report', or 'Feature Request'.

The team uses the following prompt: Classify the user feedback below.

They find that the model's classifications are inconsistent and often incorrect, frequently miscategorizing bug reports as feature requests. Analyze why this prompt is ineffective for this specific task. Then, describe two distinct, specific improvements that could be made to the prompt's description to help the model perform the classification more accurately, without providing examples of classified feedback.