Courses

Prompt engineering courses

Learn the pattern once. Write the next prompt yourself.

Short courses — ten to twenty minutes — each ending in a quiz you have to pass. They teach how a prompt works, not how one tool works, so what you learn survives the next model.

Choose a pathWhy it works
AI FLUENCY SCORE
0
0 / 36 courses

Learning paths

Pick where you want to end up. The order is done for you.

Each path runs in dependency order — no course comes before the one it builds on. A milestone counts when you pass its quiz (75%+), not when you say you're done. Move on at 85%+; between 75 and 84, redo the drills first.

FoundationsNew to prompting, or self-taught and guessing.6 milestones · 49 minFree

End state You can write a five-part prompt — role, context, task, constraints, output — that gets a usable answer on the first try, and fix one that doesn't.

  1. Prompting Mindset 101

    Notice the gap between what you want and what you ask for

    You can move on when you pass the quiz at 75%+.

    12 minFree
  2. Prompt Anatomy 101

    Name the five structural ingredients every prompt needs

    You can move on when you pass the quiz at 75%+.

    6 minFree
  3. Simple Role Prompting

    Replace generic "AI" responses with expert voice

    You can move on when you pass the quiz at 75%+.

    8 minFree
  4. Output Format Basics

    Specify formats that the model actually obeys

    You can move on when you pass the quiz at 75%+.

    9 minFree
  5. Clarity Over Cleverness

    Stop trying to be clever — be specific instead

    You can move on when you pass the quiz at 75%+.

    7 minFree
  6. Prompt Debugging 101

    Diagnose which prompt field caused a bad output (Role, Context, Task, Format, Constraint)

    You can move on when you pass the quiz at 75%+.

    7 minFree
Reliable answersYou get fluent answers that turn out to be wrong.6 milestones · 80 min2 free · rest Pro

End state You can make a model show its reasoning, bound what it may claim, and catch a wrong answer before you act on it.

  1. Basic Chain-of-Thought

    Use the phrase "think step by step" to unlock dramatically better reasoning

    You can move on when you pass the quiz at 75%+.

    10 minFree
  2. Negative Instructions

    Use negative constraints to prevent recurring failure modes

    You can move on when you pass the quiz at 75%+.

    6 minFree
  3. Constraint Calibration

    Run the constraint-stress test on any prompt in under a minute

    You can move on when you pass the quiz at 75%+.

    12 minPro
  4. Advanced Chain-of-Thought Mastery

    Spot tasks where naive CoT picks the first plausible cause and stops thinking

    You can move on when you pass the quiz at 75%+.

    22 minPro
  5. Self-Consistency Voting

    Apply majority-vote sampling to high-stakes reasoning tasks

    You can move on when you pass the quiz at 75%+.

    14 minPro
  6. Stop Hallucinations Forever

    Force the model to cite a specific source for every factual claim

    You can move on when you pass the quiz at 75%+.

    16 minPro
Prompts other people rely onYour prompts run in a product, a team workflow or a client deliverable.6 milestones · 86 min1 free · rest Pro

End state You can ship a prompt with examples, a fixed output contract and a rubric that tells you when a change made it worse.

  1. Single-shot vs Iteration

    Decide one-shot vs multi-turn in under 5 seconds

    You can move on when you pass the quiz at 75%+.

    7 minFree
  2. Few-Shot Pattern Recognition

    Pick the 2-4 examples that compress your style into a transferable pattern

    You can move on when you pass the quiz at 75%+.

    16 minPro
  3. Output Formatting Secrets

    Specify formats the model actually obeys 99% of the time (vs. 70% naive)

    You can move on when you pass the quiz at 75%+.

    18 minPro
  4. Multi-turn Refinement

    Decide when iteration outperforms one-shot

    You can move on when you pass the quiz at 75%+.

    13 minPro
  5. Prompt Evaluation Rubrics

    Author a 4-criterion rubric for any production prompt

    You can move on when you pass the quiz at 75%+.

    18 minPro
  6. Meta-Prompting

    Use the model as a prompt-engineering collaborator

    You can move on when you pass the quiz at 75%+.

    14 minPro
Agents and planningYou want the model to plan and act over several steps, not answer once.5 milestones · 106 min0 free · rest Pro

End state You can scope a multi-step job, give the model a role that holds, and build a simple tool-using agent that knows when to stop.

  1. Pre-Project Thinking Framework

    Decompose any AI project into the 4-question pre-flight (audience / decision / format / measure)

    You can move on when you pass the quiz at 75%+.

    16 minPro
  2. Role & Persona Prompting That Actually Works

    Build 4-axis personas: expertise + voice + taste + decision frame

    You can move on when you pass the quiz at 75%+.

    16 minPro
  3. Tree-of-Thoughts for Planning

    Apply ToT to problems with multiple viable strategies (entry, hiring, build-vs-buy, pricing)

    You can move on when you pass the quiz at 75%+.

    22 minPro
  4. Building Simple Agents (ReAct)

    Write tool descriptions that get selected reliably (vs. the 60%-selection default)

    You can move on when you pass the quiz at 75%+.

    24 minPro
  5. The Architect Mindset

    Internalize the Architect Mindset (Specify · Constrain · Verify) as a daily reflex

    You can move on when you pass the quiz at 75%+.

    28 minPro
FOUNDATION · FREE FOR EVERYONE

The five mindset shifts every prompter needs.

Start here. Each is under 10 minutes, taught by She, ends with a canvas you can drop into Studio.

QUICK WINS · FREE FOR EVERYONE

Eight short patterns. Five minutes each.

Bite-sized prompt patterns most working teams never learn. All free, all skippable, all immediately useful.

ADVANCED TECHNIQUES · PRO

Where deliberate practice compounds.

Self-consistency voting · few-shot pattern recognition · meta-prompting · rubric evals. The Chip-Huyen / Wei et al. canon, taught by example.

PRO · ADVANCED14 MIN
Self-Consistency Voting
Run the same prompt 3-5 times → take the majority answer. +12-18% accuracy on reasoning tasks (Wang et al. 2022).
LOCKED
PRO · ADVANCED16 MIN
Few-Shot Pattern Recognition
Show 2-4 examples → the model maps the relational structure, not the surface details (Chase & Simon's chess masters, applied).
LOCKED
PRO · ADVANCED12 MIN
Constraint Calibration
Too few constraints → drift. Too many → brittle. The 5-constraint sweet spot and how to find it.
LOCKED
PRO · ADVANCED13 MIN
Multi-turn Refinement
Conversation as compute. Use 3-5 turn loops when the goal is discovered, not pre-specified.
LOCKED
PRO · ADVANCED14 MIN
Meta-Prompting
Ask the model to write its own prompt. The technique that compresses 4 hours of prompt-engineering into 4 minutes.
LOCKED
PRO · ADVANCED18 MIN
Prompt Evaluation Rubrics
Build your own grader. The Chip-Huyen-style eval rubric that catches drift before it ships.
LOCKED
PRO · ADVANCED22 MIN
Advanced Chain-of-Thought Mastery
Stack CoT with role + self-critique for 60%+ accuracy on multi-evidence reasoning.
LOCKED
PRO · ADVANCED16 MIN
Role & Persona Prompting That Actually Works
Multi-dimensional personas that shape voice, taste, and decision-making — not a frankenstein average of 10,000 examples.
LOCKED
PRO · ADVANCED18 MIN
Output Formatting Secrets
Schema enforcement, prefilling, XML tags — the production tricks vendors don't teach.
LOCKED
PRO · ADVANCED16 MIN
Stop Hallucinations Forever
Production-grade hallucination prevention via citation-or-refuse + admit-uncertainty + claim triage.
LOCKED
PRO · ADVANCED24 MIN
Building Simple Agents (ReAct)
Thought → Action → Observation — the foundation of every modern agent, with halt conditions that don't burn $500.
LOCKED
PRO · ADVANCED22 MIN
Tree-of-Thoughts for Planning
When linear reasoning fails — enumerate branches, score independently, surface trade-offs, recommend a hybrid path.
LOCKED
ROLE SPOTLIGHTS · PRO

Real prompts for real professions.

Legal · SaaS · Founders · Marketing · Creators · Video Clippers. Pulled from 2025–2026 practitioner data, not generic templates.

MASTERY & INTERNALIZATION · PRO

Our goal is for you to not need us.

The capstone path: daily drills, the Prompt Literacy Audit, the Architect Mindset synthesis. Graduate the platform.

WHY IT WORKS
Schemas are active organizations of past reactions or past experiences. — Frederic Bartlett · 1932
Deliberate practice requires targeted reps, immediate feedback, and spaced intervals. — K. Anders Ericsson · 1993