RISEN
Role · Input · Scenario · Expectation · Nuance — The heavyweight. Built for multi-step work where the edge cases matter as much as the main path.
Two versions circulate
RISEN circulates in two forms. This site uses Role · Input · Scenario · Expectation · Nuance. You will also see Role · Instructions · Steps · End goal · Narrowing, which splits the work into explicit steps instead of describing a scenario. Neither has a vendor source; if you share a RISEN prompt, say which one you mean.
Use it for
Complex analysis, strategy, policy drafting, and anything you would brief a consultant on. Nuance is the differentiator: it is a dedicated place for the constraints everyone forgets to mention.
Avoid it for
Short tasks. Five rich slots on a simple request produces bloated, over-hedged output — and on reasoning models, heavy scaffolding can actively suppress the model’s own planning.
The slots
Each letter is a question the prompt has to answer.Role
Whose judgement?
For complex work, describe what the role is accountable for, not just its job title.
Example: A workplace-policy adviser who has to defend the final document to both a works council and a data protection officer.
Input
What material are you supplying?
The actual text, data, or facts to work from. Paste it rather than describing it.
Example: Current policy (2019): staff may use personal devices for email only. 140 employees, 40 hybrid. Two incidents last year involving chat tools.
Scenario
What situation is this happening in?
The pressures, the stakeholders, the timing, and what has already been decided.
Example: The board wants a new AI-tool policy before the works council meets in six weeks. IT wants restrictions; the commercial team is already using these tools daily.
Expectation
What does success look like?
The outcome and the shape together, since RISEN has no separate format slot.
Example: A one-page policy draft with numbered clauses, plus a short list of the decisions the board still has to make.
Nuance Optional
What are the edges and exceptions?
Sensitivities, hard limits, things that must not appear, and known disagreements. This slot is where RISEN earns its cost.
Example: Do not propose monitoring software — the works council rejected that in 2024. Do not cite specific legal articles; our counsel adds those. Assume no budget for new tooling this year.
Strengths
- Input is a real slot for source material, which most frameworks lack.
- Nuance captures the constraints that usually only surface after a bad first draft.
- Scenario gives the model a situation to reason inside rather than a bare instruction.
Weaknesses
- No format slot at all — output structure has to be smuggled into Expectation.
- Two variants circulate under the same name, so shared RISEN prompts are ambiguous.
- Slowest of the frameworks here to fill in properly.
Sources and grounding
- 1Prompting best practices — Use examples effectivelyAnthropic (Claude Platform docs)Vendor grounding for supplying input material and explicit constraints.
- 2Prompt engineering (how-to guide)Meta (Llama developer docs)Meta recommends combining role definitions, rules, restrictions and examples — the same combination RISEN encodes.