Skip to content

Saved Prompts

Saved Prompts are human-curated, org-wide prompt templates scoped to a specific AI Analyst. Once saved, any user in your workspace can run them from the Explorer homepage or by typing / in the chat input.

Only Admins can create, edit, or delete saved prompts.

Info

Looking for how to use saved prompts as a User? See Saved Prompts in Explorer.

Saving a prompt from a completed conversation

The quickest way to add a saved prompt is directly from a result you know works:

  1. Open the completed conversation in Explorer or Studio
  2. Click the command icon on the task
  3. The prompt is immediately saved to the org library for that AI Analyst
  4. A toast notification confirms the save — you can edit the title inline at this point

The title is auto-derived from the prompt text, but rename it to something descriptive (e.g. "Weekly Revenue by Region" rather than the raw question).

Info

Only the prompt is saved — not the answer. Answers go stale as your data changes. The value is the proven question that reliably produces a good result.

Managing saved prompts in Studio

To view and manage all saved prompts for an AI Analyst:

  1. Navigate to the AI Analyst in Studio
  2. Open the Saved Prompts tab (alongside Instructions, Suggestions, etc.)

From here you can:

  • View all saved prompts, with the prompt text and a link to the original task (if applicable)
  • Add a new prompt manually — without needing to save it from a conversation first
  • Edit the title or prompt text of any saved prompt
  • Delete prompts that are no longer relevant
  • Run a prompt directly from the tab to open a pre-filled conversation

Variable placeholders

Prompts can include placeholders like [region] or [timeframe]. When a user selects a prompt with placeholders — either from the homepage or via the / trigger — the placeholder text is highlighted for quick fill-in, with Tab to jump between them.

Example: Show me revenue by [region] for [timeframe]

Using saved prompts for regression testing

When you iterate on the semantic layer or agent instructions, your saved prompts become a set of "golden queries" — the canonical questions this AI Analyst must answer correctly.

After making changes to a model, metric, or guideline, run through your saved prompts from the Saved Prompts tab to verify the AI Analyst still produces the right results. You can also use the / trigger in a Studio chat session to quickly pull up any saved prompt mid-conversation.