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

# Filter Prompts by Strategy, Intent, and Output Type

> Find the right prompt fast using strategy, intent, and output-type filters — in the dashboard and inline inside the """ picker without leaving your page.

SprintBrain organises your prompt library along three independent axes — strategy, intent, and output type — so you can find the right prompt in seconds rather than scrolling through an unorganised list. The same filters are available in the dashboard and inside the inline `"""` picker, so narrowing the library never interrupts your flow.

## Filter dimensions

<AccordionGroup>
  <Accordion title="Strategy — how the prompt instructs the model to think" icon="diagram-project">
    Set the **Strategy** to signal the reasoning approach built into your prompt. Available values: **Chain of Thought (CoT)**, **Tree of Thought (ToT)**, **Few-shot**, **One-shot**, **RAG**, and **Agentic**. Filtering by strategy is especially useful when you want only prompts that use a specific reasoning pattern — for example, all CoT prompts for your Coding workflows.
  </Accordion>

  <Accordion title="Intent — what the prompt is designed to accomplish" icon="bullseye">
    Set the **Intent** to categorise the task the prompt performs. Available values: **Writing**, **Coding**, **Support**, **SEO**, **Analysis**, **Planning**, **Research**, and **Teaching**. Combining an intent filter with a strategy filter lets you find, for example, all RAG prompts tagged for Research in two clicks.
  </Accordion>

  <Accordion title="Output Type — the format the model should produce" icon="file-lines">
    Set the **Output Type** to describe the format your prompt asks the model to return. Available values: **JSON**, **Markdown**, **SOP**, and **Plain**. This filter is particularly useful when you need output in a specific machine-readable or structured format and want to rule out prompts that produce free-form text.
  </Accordion>
</AccordionGroup>

Set these fields when you create or edit a prompt. SprintBrain's intent classifier also reads your block content as you type and suggests a strategy and intent automatically — click **Apply** in the suggestion banner to accept the recommendation.

## Using filters in the dashboard

Open the **Prompts** page and use the filter bar at the top of the list to narrow by any combination of strategy, intent, and folder.

<Steps>
  <Step title="Select a Strategy filter">
    Click the **Strategy** dropdown and pick one value — for example, **CoT** — to show only Chain of Thought prompts. The active filter highlights in blue. Click it again to deselect.
  </Step>

  <Step title="Select an Intent filter">
    Click the **Intent** dropdown and pick a category — for example, **Coding** — to further narrow the list. Combine with a strategy filter to find, say, all Coding prompts that use Chain of Thought.
  </Step>

  <Step title="Filter by folder">
    Click a folder chip to scope the view to prompts in that folder. The chip row shows only folders that contain at least one prompt, so you never see an empty folder in the filter.
  </Step>

  <Step title="Clear filters">
    When one or more filters are active, a **Clear N filters** button appears. Click it to reset the view to your full library.
  </Step>
</Steps>

## The `"""` picker — filtering without leaving the page

When you type three double-quotes (`"""`) in any text field, the **prompt picker overlay** opens at your cursor. You get the same filter controls inline — without navigating away from the page you are typing on.

Use the **Strategy** and **Intent** dropdowns inside the picker to narrow the results. Once you find the right prompt, press **Enter** or click it to insert it directly at your cursor. The overlay closes and your prompt appears in the field.

<Note>
  If you have customised your prompt trigger in **Settings → Triggers**, use your configured sequence instead of `"""`.
</Note>

## Per-prompt keyboard shortcuts

For prompts you use constantly, set a shortcut so you can expand them without opening the picker at all.

<Steps>
  <Step title="Open the prompt editor">
    Click a prompt card to open the editor, or click **New prompt** to create one.
  </Step>

  <Step title="Enter a shortcut">
    In the **Shortcut** field, type a short identifier — for example, `coderev`. Use only letters, numbers, hyphens, or underscores. The field sanitises input automatically (spaces become underscores).
  </Step>

  <Step title="Save">
    Click **Save changes** (or **Create prompt**). The shortcut is now active across every device where you use SprintBrain.
  </Step>

  <Step title="Expand in any field">
    In any text input, type your prompt trigger prefix + your shortcut + **Tab**. With the default prefix, that is `/coderev` + **Tab**. The full prompt expands at your cursor immediately.
  </Step>
</Steps>

<Tip>
  Tag every prompt with both an intent and an output type when you create it — even if you do not plan to filter right away. A well-tagged library makes the inline `"""` picker dramatically faster: two filter clicks and you are looking at exactly the prompts that fit your current task, rather than scanning by name. Consistent tagging also gives the quality scorer a more meaningful comparison when it shows how your prompt ranks against others in the same category.
</Tip>


## Related topics

- [AI Prompt Library: Store and Expand Prompts Anywhere](/prompts/overview.md)
- [SprintBrain: Free AI Prompt Library and Text Expander](/introduction.md)
- [Prompt Quality Score: Grade and Improve Your Prompts](/prompts/quality-score.md)
- [SprintBrain Quickstart: Install, Sign In, and Expand](/quickstart.md)
- [Create and Structure AI Prompts with Prompt Blocks](/prompts/create-prompts.md)