Prompt Pages in XTM One
This page complements Prompts in XTM One.
That guide explains when prompts are useful. This page explains what the prompt pages show and how to use them.
What you see in the sidebar
In the left navigation, this area appears as Prompts.
The page title in the main content area is also Prompts.
Prompts list page
The list page is a reusable library view.
Use this page to:
- search prompt templates by name or by a remembered wording fragment when you know the kind of request you want but not the exact prompt title
- filter by visibility to distinguish your own prompts from shared or company-managed prompts
- filter by tags to narrow the library to a topic, team, or business workflow
- compare prompt cards before opening one, especially when several prompts appear similar
- check whether a reusable prompt already exists before creating another version
- create a new prompt when you have permission and want a repeatable instruction that others can reuse
The header contains:
- page title
Prompts - short explanation that prompts are reusable templates for agents and chat conversations
Add Prompt
The filter row contains:
- search field
- visibility filter
- tag filter
- total count on the right
Each prompt card shows:
- prompt name
- short description
- content preview snippet
Company-managedorSharedstate when relevant- tags
That card design is useful because it lets you judge quickly:
- whether the prompt looks relevant
- whether it is a broadly shared asset or a more local one
- whether the content style feels close to what you need before opening the detail page
Empty states on the list page
If no prompts exist yet, the page shows an empty state encouraging the user to create the first reusable prompt template.
If prompts exist but the filter removes them all, the page shows No prompts matching your filters.
Prompt detail header
When you open a prompt, the header shows:
- back button to
Prompts - prompt name
- the label
Prompt Template - content length in characters
Company-managed,Shared, andRead-onlystates when relevant
The main destructive action in the header is delete. If the user cannot manage the prompt, that action is not available.
Tabs on the detail page
The prompt detail page is organized into:
OverviewConfigurationActivity
Overview tab
The Overview tab combines summary metrics with readable content preview.
The KPI row shows:
- number of agents using the prompt
- content length
- assignment run count from agents using the prompt
- last activity
Those indicators help the user answer practical questions such as:
- is this prompt just a draft or already reused in production-like workflows
- is it a short reusable instruction or a longer structured template
- if I change it, how many downstream uses might I affect
The main overview card shows:
- description
- content preview
- tags
- created and updated timestamps
At the bottom, the user sees an Agents Using card. This is useful when someone wants to understand the blast radius of editing a shared prompt.
Configuration tab
The Configuration tab uses a standard read mode and edit mode.
In read mode, you review:
- name
- description
- full content
In edit mode, you can change:
- prompt name
- tags
- description
- content
Hide from chat- sharing and visibility settings
- public catalog state when that capability is available
The most important distinction is this:
- editing the content changes what the prompt says
Hide from chatchanges where it is surfaced in the UI- sharing settings change who can reuse it
The content editor includes a variable insert control. That means you can insert platform variables directly from the UI instead of memorizing the syntax.
Hide from chat changes whether the prompt appears in the chat prompt picker. A hidden prompt can still remain useful for agent configuration.
Activity tab
The Activity tab helps the user answer operational questions such as:
- who changed the prompt
- when it changed
- whether a recent run problem started after a prompt edit
This is particularly useful for shared prompts, where several users may rely on the same text asset.
Read-only and shared states
Prompt pages can behave differently depending on ownership:
- private prompts can usually be edited by their owner
- shared prompts are visible in a group-sharing context
- company-managed prompts are visible to everyone but read-only for non-admins
When a prompt is read-only, you can still inspect the content and usage. You just cannot switch into edit mode.
What the prompt UX is good at
The prompt UX is optimized for:
- quick scanning from the list page
- easy reading of content before reuse
- safe editing with clear ownership context
- understanding downstream impact through
Agents Using
Good habits on these pages
- Read the content preview before creating a duplicate prompt.
- Check
Agents Usingbefore editing a shared prompt. - Use
Hide from chatwhen the prompt is operational and not meant for casual use. - Keep names and descriptions clear enough that other teams can reuse the prompt without guessing.