Framework Quick-Reference Cards

Print These. Pin Them. Use Them Daily.

How to Use These Cards

Each card is a standalone reference for one framework or technique. Print them, screenshot them, or pin them next to your monitor. They’re designed to be useful without re-reading the full module.

Card 1: The 3 Questions (Everyday Default)

Before you type any prompt, answer:

# Question What It Does
1 What do I want? Tells the AI your intent — the #1 factor
2 Who am I? Sets your expertise level and context
3 How will I use this? Changes tone, depth, and focus

Example: > “I’m a first-year teacher at a Title I school [who am I]. I need 3 warm-up activities for 7th graders learning fractions [what I want]. These go in my lesson plan for Monday’s observation [how I’ll use this].”

When to use: Every single AI interaction. Takes 10 seconds. Works every time.

Card 2: BRIEFS (Precision First Message)

When stakes are high or you’ve been getting generic results, add precision:

Letter Means Ask Yourself
B Background Who am I?
R Role What’s my expertise level?
I Intent What do I actually want? (MOST IMPORTANT)
E Execution What constraints matter?
F Format How long? What structure?
S Stakes What will I do with this output?

The shortcut: Intent is the anchor. Start there. Add others as needed.

When to use: Loan applications, client deliverables, medical questions, important emails, any task where “close enough” isn’t enough.

(Source: Yatawara, 2026 — LLM Whisperers Study, 26/26 model convergence)

Card 2b: The BRIEFS Production Ladder

How to become fluent in BRIEFS, not just aware of it. Follow the ladder in order — each rung reduces scaffolding by one step:

Rung What You Do Why This Rung
1 Diagnose — read 6 weak prompts, name the missing BRIEFS element You can’t write what you can’t spot
2 Insert — add one BRIEFS element at a time to a bare prompt Feel each element as a distinct move
3 Complete — fill 3 partial BRIEFS skeletons First production with guardrails
4 Produce — write 3 full BRIEFS prompts from scratch Full production, low stakes
5 Handshake — practice the restate-and-confirm exchange The phase most users skip
6 Select technique — pick zero-shot / few-shot / CoT / role / structured Technique ≠ BRIEFS — separate decision
7 SMARTS — write all 6 turns on one real complex task Multi-turn fluency
8 Anti-pattern — diagnose 7 broken prompts, one-move repair Pattern recognition for failure
9 Transfer — rewrite a prompt across 3 domains; red-team where BRIEFS fails The true test of mastery
10 Peer critique — score a partner’s prompt with the rubric The reviewer learns more than the reviewee

When to use: The first time you teach yourself or anyone else BRIEFS. Once a rung feels automatic, move up. Don’t skip rungs — cognitive load research (Renkl & Atkinson, 2003) is clear that novices who skip scaffolding end up worse than those who took the time.

(Source: Module 4 Exercise Ladder, built on the faded-worked-examples evidence base.)

Card 3: SMARTS (Multi-Turn Reasoning)

For complex tasks where you need to THINK with AI, not just get an answer:

Step What You Do What You Say
S — State Clarify the question “Can you rephrase this so I understand the task?”
M — Map Plan your approach “Here’s my plan: [steps]. Is this sound?”
A — Ask Gather resources “What concepts do I need? Remind me of the key terms.”
R — Reason Work through it “Check this step before I continue.”
T — Think Get unstuck “Give me a hint, not the answer.”
S — Summarize Confirm what you learned “Summarize what we did and what I should remember.”

When to use: Studying for exams, analyzing business decisions, learning new concepts, solving multi-step problems, any task where the process matters as much as the answer.

(Source: Yatawara, 2025 — ELEVATE Project)

Card 4: The Complete Protocol (BRIEFS + SMARTS)

BRIEFS → Design your first message
    ↓
CALIBRATION HANDSHAKE → AI restates, you confirm
    ↓
SMARTS → Guide the reasoning loop
    ↓
VERIFIED OUTPUT → Check, iterate, done
Situation Use
Simple one-shot task BRIEFS only
Quick familiar question Neither — just ask
Complex reasoning task BRIEFS to open + SMARTS to guide
Learning something new SMARTS only
High-stakes output Full protocol + extra verification

Card 5: The Critique Rubric (For Peer Review)

Score any BRIEFS prompt 0 / 1 / 2 on five dimensions. Ten points maximum.

Dimension 0 1 2
Intent specified Topic only General goal Specific outcome + use
Context right-sized Missing or drowning Adequate Exactly what the model needs, no filler
Execution constraints None Soft (“be concise”) Hard and specific (“≤300 words”)
Format stated None Vague (“bullet points”) Unambiguous structure
Stakes transparent Implied Named Named + consequence named

Interpretation: - 8–10 = production-ready - 5–7 = one calibration handshake will fix it - 0–4 = rewrite before sending

The critique script (one targeted fix, not five vague ones): 1. Say what’s working first (one sentence). 2. Identify the weakest dimension by score (one sentence). 3. Propose the one revision that would raise that score to 2 (one sentence). 4. Stop. One targeted fix beats five vague suggestions.

(Source: Module 4, Exercise 4.J — adapted from software-engineering code review practice.)

Card 6: The 4-Step Verification Protocol (Module 5)

Before trusting any AI-generated claim:

Step Action Tool
1 Check the source — Does the cited source actually exist? Google Scholar, library databases
2 Read the source — Does it actually say what the AI claims? Click the link. Read the passage.
3 Cross-reference — Does a second independent source confirm? Different search, different database
4 Check the date — Is this still current? Look for more recent publications

Red flags that demand verification: Specific statistics, named studies, direct quotes, medical claims, legal claims, anything with a number attached.

Card 7: The 5-Filter Hype Test (Module 13)

When someone claims “AI will change everything” or “AI can’t do X”:

Filter Question
1 Who’s claiming this? Researcher? Company selling something? Journalist?
2 Where’s the evidence? Peer-reviewed? Demo? Just a press release?
3 What’s the incentive? Who profits from you believing this?
4 Does it match the trajectory? Incremental improvement or magic leap?
5 What do credible skeptics say? Not doomers — informed critics.

Rule of thumb: We overestimate AI in the short term, underestimate it in the long term. Plan for both.

Card 8: AI Ethics Decision Framework (Module 10)

Before deploying AI in any context:

Question Why It Matters
Who benefits? AI often benefits the organization more than the people affected
Who could be harmed? Harms fall disproportionately on marginalized communities
What data goes in? Never enter PII, proprietary info, or student data without policy
Is there a human in the loop? High-stakes decisions need human review
Would I be comfortable if this were public? Act as if your AI use will be disclosed
Am I replacing judgment or augmenting it? AI supports decisions; humans make them

Card 9: Prompt Patterns (Universal)

Pattern Template When
Role + Task “You are a [role]. [Task].” Domain expertise needed
Step-by-Step “[Task]. Think step by step.” Accuracy matters
Examples First “Here are examples: [X]. Now: [Y].” Specific format needed
Constraints “[Task]. Do NOT [X]. Under [limit].” Need boundaries
Iterate “Good start. Make it [adjustment].” Refining output
Meta-Prompt “What info do you need from me?” You’re stuck
Challenge “What’s wrong with this?” Want honest feedback
Critic “Don’t rewrite. Diagnose problems.” Reviewing your draft

Print these. The best framework is the one you actually use.

Back to Modules