# All Prompts (https://app.infrasity.com/docs/geo-dashboard/all-prompts)





All Prompts is the detailed, prompt-level view behind your [Overview](/geo-dashboard/overview) citation rate. Instead of a single aggregate percentage, it shows exactly how your site ranks for every individual user question being tracked, across ChatGPT, Gemini, Claude, and Perplexity, organized so you can immediately tell what's working, what's close, and what's missing entirely.

<Screenshot src="/images/geo-all-prompts.png" alt="All Prompts table showing tracked prompt counts across clusters, with Total/Winning/Needs Work/Not Cited counts and a per-model ranking table" caption="All Prompts: every tracked prompt, its score, cluster, and per-model ranking." />

## Header [#header]

* **Prompt / cluster count and last-scanned timestamp**: the total tracked prompt count, how many clusters they're grouped into, and when this data was last refreshed.
* **+ Add Prompt**: manually add a new prompt to track, outside of the automated theme-generation flow described on the Overview page.

## Summary counts [#summary-counts]

Four stat tiles break the full prompt set down by outcome:

* **Total**: every prompt being tracked.
* **Winning**: prompts where your site is cited.
* **Needs Work**: prompts with partial or weak citation performance.
* **Not Cited**: prompts where your site doesn't appear at all.

<Callout type="gotcha" title="Categories aren't a documented formula">
  The exact rule for what separates "Winning" from "Needs Work" (score threshold, number of models citing you, etc.) isn't shown on this page. Treat the three categories as a rough triage, not a precisely documented scoring formula.
</Callout>

## Filters and search [#filters-and-search]

* **All / Winning / Opportunities tabs**: quickly filter the table to just the prompts in each bucket, with a live count shown on each tab.
* **Search prompts…**: free-text search across prompt wording.
* **All Clusters dropdown**: narrow the table to a single topic cluster.
* **Group**: presumably groups table rows by cluster instead of a flat list; behavior isn't demonstrated further here.
* **Cluster pills**: a row of quick-filter chips for individual clusters, letting you jump to one cluster without opening the dropdown.

## The table [#the-table]

Each row is one tracked prompt:

* **# / Score**: row number and a circular score badge summarizing citation strength for that prompt across all models.
* **Prompt**: the actual user question being evaluated.
* **Cluster**: which topic cluster the prompt belongs to.
* **Gemini / Perplexity / Claude / ChatGPT**: one column per AI model, showing your citation rank for that prompt on that model (`#1`, `#2`, …) or `—` if you weren't cited by that model for that prompt.

<Callout type="info" title="Rank, not raw score, per model">
  The per-model columns show *where you ranked* among cited sources (`#1`, `#2`, etc.), not a percentage: a prompt can rank on one model and show `—` on another in the same row, which is exactly the kind of gap the [Overview](/geo-dashboard/overview) page's per-model citation rates are summarizing in aggregate.
</Callout>

<RelatedPages
  pages="[
{ title: &#x22;Overview&#x22;, description: &#x22;Monitor your AI visibility and citation performance across LLMs.&#x22;, href: &#x22;/geo-dashboard/overview&#x22; },
{ title: &#x22;Clusters&#x22;, description: &#x22;Group related prompts into clusters for theme-level analysis.&#x22;, href: &#x22;/geo-dashboard/clusters&#x22; },
{ title: &#x22;AEO Audit&#x22;, description: &#x22;Audit your content and docs for AI readiness.&#x22;, href: &#x22;/geo-dashboard/aeo-audit&#x22; },
{ title: &#x22;Competitors&#x22;, description: &#x22;Track competitor visibility in AI-generated responses.&#x22;, href: &#x22;/geo-dashboard/competitors&#x22; },
]"
/>
