The prompts rollup

The prompts rollup at the bottom of the reports page condenses every tracked prompt into one row, so you can see which questions work for you and which do not.

What each row shows

1. The prompt text and its topic
2. Executions: how often it ran in the window
3. Per-platform breakdown: cited or mentioned on each platform, with run counts
4. Status: whether the prompt currently ends in a citation or a mention
5. Drift: whether the answer stays stable, moves moderately, or is volatile between runs

Each row expands into recent executions with a response preview, and every execution links to its full result page with citations, mentions and fan-out detail.

How to use it

1. Sort for your best prompts: these anchor your reporting, because they are the questions customers ask before buying
2. Find the worst prompts and take them to the gap diagnosis: a prompt that runs often and never cites you is a targeted fix, not a mystery
3. Watch volatile prompts: instability makes them unreliable anchors; check what the AI now cites before drawing conclusions

The table respects the range and topic filters and exports as CSV.