> ## Documentation Index
> Fetch the complete documentation index at: https://docs.usecliro.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Key concepts

> The core vocabulary of Cliro — brands, topics, prompts, responses, citations and how they fit together.

Understanding a few core terms makes the rest of Cliro easy to navigate.

## The big picture

<Steps>
  <Step title="Your main brand + competitors">
    You track one **main brand** (your website) alongside **competitor brands**.
  </Step>

  <Step title="Topics group prompts">
    **Topics** organize your **prompts** — the questions Cliro asks the AI models.
  </Step>

  <Step title="Prompts produce responses">
    Each prompt runs against the enabled **models** and returns a **response**.
  </Step>

  <Step title="Responses become insights">
    Every response is analyzed for **brand mentions** (ranking + sentiment) and **citations** (source domains), which power your dashboards.
  </Step>
</Steps>

## Glossary

<AccordionGroup>
  <Accordion title="Brand & main brand" icon="building">
    A **brand** is a company or product you track. Your **main brand** is your own website — the one Cliro centers its analysis on. Other brands you add are treated as competitors.
  </Accordion>

  <Accordion title="Topic" icon="folder">
    A **topic** is a theme that groups related prompts — for example "project management tools" or "best CRMs for startups". Topics keep your prompts organized and your reporting readable.
  </Accordion>

  <Accordion title="Prompt" icon="message-square">
    A **prompt** is a question or query Cliro sends to the AI models, such as *"What are the best AI visibility tools?"*. Cliro tracks whether your brand appears in the answer, where it ranks, and how it's described.
  </Accordion>

  <Accordion title="Model & provider" icon="cpu">
    A **model** is a specific AI engine (e.g. ChatGPT, Claude, Gemini), grouped under a **provider** (OpenAI, Anthropic, Google). You choose which models to run your prompts against. Each model can answer the same prompt differently.
  </Accordion>

  <Accordion title="Response" icon="reply">
    A **response** is what a model returned for a given prompt at a given time. Cliro stores responses so you can see how answers — and your visibility — change over time.
  </Accordion>

  <Accordion title="Brand mention" icon="at-sign">
    A **brand mention** is a reference to a brand inside a response. Cliro records each mention's **position/ranking** and **sentiment** (positive, neutral or negative) so you can measure not just *if* you appear but *how well*.
  </Accordion>

  <Accordion title="Citation" icon="quote">
    A **citation** is a source URL or domain a model used to build its answer. Citations reveal which websites influence AI answers about your space — and where you might want to earn a mention.
  </Accordion>

  <Accordion title="Tag" icon="tag">
    A **tag** is a label you attach to prompts to slice and filter your data — for example by funnel stage, region or campaign.
  </Accordion>

  <Accordion title="Tenant & subscription" icon="users">
    Your **tenant** is your account workspace (an agency, marketing team or individual). It holds your brand, prompts and data, and is linked to a **subscription** that defines your plan limits.
  </Accordion>
</AccordionGroup>

## How prompts run

Cliro runs prompts in three ways:

| Mode                | When it happens                                          |
| ------------------- | -------------------------------------------------------- |
| **Scheduled**       | Active prompts re-run automatically on a daily cycle.    |
| **Execute now**     | You trigger a prompt on demand for instant results.      |
| **First responses** | A one-time run during onboarding so you start with data. |

<Note>
  Only **active** prompts count toward your plan's prompt limit and run on the schedule. Suggested prompts don't count until you activate them.
</Note>

## Next

<Card title="Onboarding walkthrough" icon="list-checks" href="/en/getting-started/onboarding">
  See exactly what happens in each onboarding step.
</Card>