PromptLab Docs
DocsTools
Back to App

Cases

Cases are pre-built real-world scenarios that demonstrate specific prompt engineering strategies in action. Each case has a built-in system prompt, a strategy explanation, and an interactive chat.

Overview

The Cases page bridges theory and practice. Instead of staring at a blank Playground, you start from a real-world scenario — complete with a purpose-built system prompt — and immediately see how prompt engineering decisions affect the output.

Each scenario demonstrates a specific PE strategy: Socratic questioning, empathetic persona, constraint-based planning, structured extraction, and more.

Available scenarios

ScenarioRolePE Strategy
AI TutorPatient educatorSocratic Method — guide with questions, not answers
Customer SupportEmpathetic agentEmpathy + Classification + Resolution framework
Travel PlannerExpert plannerConstraint-aware generation with budget limits
Resume AnalyzerSenior recruiterStructured gap analysis with role matching
Medical ScribeClinical documenterProfessional nomenclature + safety disclaimer injection
Code ReviewerPrincipal engineerMulti-dimension review (security + performance)

Strategy cards

Each scenario displays a Prompt Strategy card in the right panel that explains the specific technique being applied. Reading these cards is how you connect the output you see to the underlying prompting decision that produced it.

For example, the AI Tutor card explains why the system prompt never gives direct answers — it's the Socratic Method in practice.

Tip
After running a scenario, open the system prompt in the Cases panel and modify it. See what happens when you remove the empathy instruction, or add "be concise." This is how you develop prompt intuition.

How to use

  1. Select a scenario from the left sidebar
  2. Read the Prompt Strategy card to understand the technique
  3. Type a user message in the input (e.g., "I don't understand recursion" for the AI Tutor)
  4. Click Run and observe how the system prompt shapes the response
  5. Try different inputs to explore the scenario further