Overview
How might we design conversation templates so people can trust and act on AI output?
When this pattern fits
- Ideal for power users, professionals, and teams who have established workflows and want to standardize their AI interactions.
When to skip or lighten it
- Completely open-ended chat with no repeating workflows.
- Templates that go stale faster than teams will maintain them.
- Compliance contexts where every chat must start blank for audit reasons.
States
State model coming soon
Key UX elements
Key UX elements coming soon
Anti-patterns to avoid
Templates that hide the underlying prompt from the user.
No way to fork and customize a template.
Marketplace templates that request tools without disclosure.
Overwriting a shared template without version notes.
How products use it
| Product | Implementation |
|---|---|
| ChatGPT | Custom GPTs and starter prompts as reusable setups. |
| Claude | Projects and prompt starters for repeat workflows. |
| Notion AI | Saved prompts and templates for recurring docs. |
| Jasper | Template galleries for marketing chat flows. |
Implementation
Copy this prompt to generate a production-ready implementation in Cursor, Claude Code, Lovable, or any AI coding agent.
Generate a production-ready implementation of the "Conversation Templates" AI interface design pattern.
Pattern Definition:Frequently asked questions
What belongs in a conversation template?
System or role instructions, starter user prompt, default tools, and optional example outputs—visible to the person launching it.
How do templates differ from prompt libraries?
Prompt libraries are copy-paste strings. Conversation templates launch a full chat configuration, sometimes with tools and memory scope.
Who can edit shared templates?
Owners and admins; everyone else should fork. Show last updated and publisher.
When should a product suggest creating a template?
After a user repeats a similar opener or stars a successful thread.