RTF Framework
Role · Task · Format
Minimum viable prompt structure. Three fields that cover the three most impactful variables in any prompt.
- Definition
- The RTF framework (Role · Task · Format) is a prompt engineering structure that breaks your AI request into 3 discrete fields. It is best suited for professional tasks where a role matters.
The 3 Fields
Role
The expert persona the AI should adopt — what specialist knowledge to draw on.
Task
Precisely what you want the AI to do.
Format
How the output should be structured and presented.
Real Example
Scenario: Creating a training module on prompt engineering
Role: Instructional designer with 10 years of corporate training experience. Task: Create a 30-minute training module on prompt engineering fundamentals for non-technical employees. Format: Module outline with 5 sections, learning objectives, and a quiz at the end.
When to Use RTF
- ✓Professional tasks where a role matters
- ✓When you want structured output without verbose setup
- ✓Training and educational content generation
- ✓When APE is too minimal but full frameworks are overkill
- ✗Tasks requiring audience definition (use CRAFT)
- ✗Tasks needing tone/style separation (use CO-STAR)
- ✗Complex multi-step workflows (use RISEN)
Frequently Asked Questions
What does RTF stand for?
RTF stands for Role, Task, and Format — a minimal three-field framework covering the three highest-impact variables in any prompt.
When should I use RTF instead of APE?
Use RTF when assigning an expert role improves output quality. Use APE when the task is simple enough that a role would add no value.
Is RTF good for generating training content?
Yes. RTF works especially well for training and educational content because the Role field sets expert authority and the Format field ensures consistent structure.