CRIT Prompt Framework: Complete Guide with Examples (2026)

Last updated June 28, 2026 · 2 examples · Works with ChatGPT, Claude, Gemini
Quick Answer

CRIT is a four-part prompt framework: Context sets the background, Role assigns the AI a persona, Interview instructs the AI to ask clarifying questions before proceeding, and Task defines what to produce after the interview. The Interview field is unique — it makes CRIT the only framework that actively gathers missing information before executing.

What Is the CRIT Prompt Framework?

CRIT's defining characteristic is the Interview field. Rather than forcing you to anticipate every piece of context the AI might need, CRIT instructs the AI to ask targeted clarifying questions first. This makes it particularly valuable for open-ended, creative, or advisory tasks.

What Does CRIT Stand For?

Interview-style framework — AI gathers context before executing.

C
Context
The situation, background, or problem to be solved.
R
Role
The expert persona the AI should adopt. Determines the type of questions it asks.
I
Interview
Instructions for the AI to ask clarifying questions before executing. Specify how many questions (typically 3–5) and what areas to probe.
T
Task
What the AI should produce after the interview is complete.
+R
Rules (optional)
Cross-cutting constraints added by Promptary as an optional fifth field.

When to Use CRIT

Use CRIT for open-ended creative tasks, strategic advice, consulting-style outputs, or any task where quality depends heavily on details you might not know to provide upfront. If you always know exactly what you want, RACE is faster.

CRIT Examples

Brand strategy
Context: I am launching a new developer tool and need to define my brand positioning and messaging strategy.
Role: You are a brand strategist with 15 years of experience positioning B2B developer tools.
Interview: Before writing the strategy, ask me up to 5 targeted questions to understand: my target customer, the core problem I solve, existing alternatives, my key differentiator, and any brand constraints.
Task: After the interview, write a one-page brand positioning document with: positioning statement, 3 key messages, and tone of voice guidelines.
Career advice
Context: I am a software engineer considering a transition to product management.
Role: You are an executive career coach who has helped 200+ engineers move into product roles.
Interview: Ask me 4 questions to understand my current situation, motivations, timeline, and risk tolerance before giving advice.
Task: After the interview, give me a structured assessment and a 90-day action plan if I decide to proceed.

CRIT vs Other Frameworks

CRIT vs RACE: RACE assumes you know all the context needed. CRIT is for when you do not — the AI helps gather it. CRIT is slower (requires a conversation turn) but produces better results on open-ended tasks.

Use this CRIT template in Promptary — free

Save your first CRIT prompt in the structured editor and get a permanent REST API endpoint. Personal plan is free forever, no credit card required.

Start free

Frequently Asked Questions

What does CRIT stand for in prompt engineering?

CRIT stands for Context, Role, Interview, and Task. The Interview field instructs the AI to ask clarifying questions before executing the task, rather than attempting it with incomplete information.

Does CRIT require multiple conversation turns?

Yes. CRIT is designed for conversational use: you send the prompt, the AI asks questions, you answer, then the AI executes. It is not suited for single-shot automation.

How many interview questions should I specify?

3–5 is the sweet spot. Specify the areas to probe, not the exact questions, and let the AI formulate them based on its role.