RTF vs RCTF vs RCTFVV: which prompt framework should you use?
A side-by-side comparison of the RTF, RCTF and RCTFVV prompt frameworks, with the same task written three ways and simple rules for choosing between them.
By the ThinkWithPrompt editorial team at Synottic · Updated 11 October 2026
RTF, RCTF and RCTFVV are three levels of the same idea. Each adds structure to the one before it, and each extra part is worth including only when the task needs it. Here is how they differ and how to choose.
The three frameworks at a glance
| Framework | Parts | Best for | Main risk if you use it for everything |
|---|---|---|---|
| RTF | Role, Task, Format | Quick, everyday tasks with little background | Generic output when the situation actually matters |
| RCTF | Role, Context, Task, Format | Work for a specific audience or situation | Slightly more effort to write |
| RCTFVV | RCTF plus Verification and Validation | High-stakes work someone will rely on | Longer answers; overkill for simple tasks |
The same task, three ways
RTF, enough for a routine message:
RCTF, when the audience and situation change what “good” means:
RCTFVV, when a mistake would be costly:
How to choose
- If you could explain the task to a colleague in one sentence with no background, use RTF.
- If the right answer depends on who it is for or what has happened, use RCTF. This is the right default for most work.
- If the output will be relied on by others, sent externally, or touches money, law, health or people decisions, use RCTFVV.
Why the verification steps help
Language models are fluent even when they are wrong. Asking for a self-check against your stated requirements, and for a list of assumptions, does not make the model infallible, but it moves likely problems to where you can see them. You still need to review the result; the framework makes that review faster.
Where other frameworks fit
These three describe the sections of a prompt. Techniques such as few-shot prompting (adding worked examples) and chain-of-thought prompting (asking for step-by-step reasoning) can be combined with any of them. See our guide to few-shot vs chain-of-thought.
Frequently asked questions
What is the difference between RTF and RCTF?
RTF has three parts (Role, Task, Format); RCTF adds a Context section between the role and the task. Use RTF for quick everyday requests and RCTF whenever the right answer depends on the audience, the situation or constraints the model can't know.
When should I use RCTFVV?
Use RCTFVV for work someone will rely on: reports, policies, client-facing documents and anything involving money, law, health or people decisions. The extra Verification and Validation steps make the model check its draft against your requirements and list the assumptions you should confirm.
Example RCTFVV prompts from the library
- Write personalized cold outreach that gets repliesA short, relevant multi-touch outreach sequence — email and LinkedIn.
- Research a prospect before a sales conversationAn account brief: business context, priorities, likely pains, stakeholders, and conversation angles.
- Define my ideal customer profile (ICP)Defines the accounts and buyers most likely to buy, succeed, and stay.
- Draft a sales proposal or business case for a buyerA proposal built around the buyer's problem, solution, value, pricing, and next steps.
- Review pipeline health and forecastAssesses deal quality, risks, and forecast confidence, with actions per deal.
- Create a competitive battlecardHow to win against a specific competitor: positioning, traps, objections, and proof.
More guides
- How to write a good AI prompt: a practical guide
- The CO-STAR prompt framework explained, and how it compares to RCTF
- Few-shot vs chain-of-thought prompting: when to use each
- Prompt meaning: what is a prompt in AI? Parts and examples
- Prompt meaning in Hindi: प्रॉम्प्ट का मतलब, AI प्रॉम्प्ट क्या है और उदाहरण
- What is prompt engineering? A practical beginner’s guide
- What is a system prompt? How to write one, with examples
- Prompt templates: how to build reusable AI prompts, with examples
- How to prompt reasoning models: what changes with “thinking” AI
- Prompt chaining: how to break big AI tasks into steps