Improving prompt handling and rerendering logic

I am writing to request improvements to the UI/UX regarding prompt management and credit consumption. Currently, the system's behavior leads to significant accidental loss of credits.

1. Auto-Clear Prompt and Prevention of Accidental Re-execution
Currently, when I press the generate button, the prompt remains in the input field. If I accidentally press Enter or refresh the page, the same prompt is re-executed, consuming credits for the exact same image. Furthermore, even after closing the window, returning to the site sometimes triggers an automatic re-execution of the previous prompt.
I know this save time for executing the same prompt again, but as mentioned in No.2 below, 90% of the images are similar, and more of the time user wont be using same prompt again and again, compared to trying out other text-to-image. Everytime i have to delete all the words in the prompt bar and this’s kinda stressful.

• Request: Please implement a feature to clear the prompt field after execution, or add a confirmation dialog to prevent unintentional re-generation that consumes credits.

2. Improved Variation Logic for "Regenerate" (Model: Nano Banana 2)
The current "Regenerate" function produces images that are about 90% identical to the previous one. This feels like a waste of credits.

• Request: When using the regenerate/variation feature, the AI should be smart enough to provide distinct variations rather than near-identical copies.

3. Implementation of a "Copy Prompt" Button
Longer prompts are currently truncated with "..." in the UI, making it impossible to copy the full text manually. Since the "Regenerate" function (as mentioned in point 2) does not provide enough variation, I often need to copy the prompt, tweak it manually, and run it as a new generation.

• Request: Please add a dedicated "Copy Prompt" button so users can easily reuse and modify their prompts without them being hidden by the UI.

Thank you :)