Crystal can **keep memory of your previous questions**, i.e. the **context** of your queries.
This capability allows you to **ask follow-up questions** on a Topic *without having to write full sentences again*.
In particular, Crystal can store memory of all information related the **Topic** itself.
## Memory of Topic Elements
Once you ask about a **Topic**, Crystal remembers your question as part of a **conversational context**.
In particular, Crystal is able to store memory of the [**elements forming a Topic**](/fundamentals/crystal-topics.md), so that you can **ask for other elements** without re-writing the ones you first asked about.
In other words, the context memory enables you to **ask about Topics by requesting a sort of element "switch" within the Topic in context**.
{% hint style="warning" %} **Remember**
Even when asking about Topics with this modality, Crystal will only return Topic configured by the Admin in Console.
{% endhint %}
### Objective Memory (Entity Switch)
Crystal can remember the Topic's Objective as the conversational context.
This way, you can ask about a new Topic by just switching the Entities.
How? Simply ask *“And for…?”*, then add the Entities in coherence with the current Objective.
Crystal will show you a Topic that has the same Objective but different Entities!
{% hint style="success" %} **Examples**
* *(Breakdown) "And of Sales by Agent?"*
* *(Comparison) "And of Sales vs Target?"*
* *(Overtime) "And of Country?"*
{% endhint %}
Same Objective, Different Entities
### Entity Memory (Objective Switch)
Crystal can remember the Topic's Entities as the conversational context.
This way, you can ask about a new Topic by just switching the Objective.
How? Simply ask *“And the…?”*, then add an Objective (or Visualization) in coherence with the current Entities.
Crystal will show you a Topic that has the same Entities but different Objective!
{% hint style="success" %} **Examples**
* *"And the Comparison?"*
* *"And the Breakdown?"*
* *"And the Overtime?"*
{% endhint %}
Same Entities, Different Objective
### Topic Memory (Filter Switch)
Crystal can remember both Objective and Entities together (i.e. a Topic) as the conversational context.
This way, you can narrow down the same Topic with Filters, without asking for the Objective and Entity again.
How? Simply ask *“And for…?”*, then add a Dynamic and / or a Time Filter.
Crystal will show you the same Topic filtered with new Filters!
{% hint style="success" %} **Examples**
* *"And for country Italy?"*
* *"And for Italy?"*
* *"And for last quarter?"*
* *"And in 2022?"*
* *"And for Spain in 2021?"*
{% endhint %}
Same Topic, Different Filters
{% hint style="info" %} **Please Note**
In this context, if you keep asking Filters with the "and for" formulation, Crystal will *add* the new requested Filter to the Topic, keeping in memory all the Filters applied. This means that Filters already present in the Topic will remain afterwards, and the new Topic will contain both Filters.
For example, asking "and for Spain?" in the example above will give the same Topic filtered for both Italy *and* Spain.
{% endhint %}
### Filter Memory (Topic Switch)
If the previously asked Topic contained Filters, Crystal can remember them too as the conversational context.
This way, you can ask about another Topic (with different Objectives and Entities) that contains the same Filters, instead of asking for the new Topic and specifying Filters again.
How? Simply ask *“And the…?”,* then add your Topic.
Crystal will show you the new Topic filtered with the same Filters!
{% hint style="success" %} **Examples**
* *"And the List of Country Managers?"*
* *"And the Breakdown of total Sales by Country?"*
* *"And the Comparison of Sales vs Target?"*
{% endhint %}
Same Filters, Different Topics
Keep talking to your data!
***
---
# Agent Instructions: Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://docs.igenius.ai/crystal-advisor/talk-to-your-data/conversation/request-topics/context-memory.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.