Free · Reference & Training Tool

Explore the structure
of effective AI prompts
for QA work.

See how purpose, role, coverage, and guards combine to shape an AI prompt. Explore the options, understand what each selector does — then bring that structure to your own AI agent.

Works with ChatGPT Claude Gemini + any AI
composer — gherkin preset
INPUT
TEST-4821 · Dealer cannot
save vehicle if VIN already
exists in another lot

When a dealer adds a vehicle
with a duplicate VIN, save
button spins indefinitely.
No error message is shown.

AC:
1. Show inline error
2. Save button re-enables
3. Duplicate VINs blocked
OUTPUT
Feature: Duplicate VIN
  validation [TEST-4821]

Scenario: Save with
  duplicate VIN
  Given dealer on Add Vehicle
  When VIN "XYZ" in "North"
  Then inline error appears
  And Save re-enables

⚠ AMBIGUITY: "re-enable"
  timing unclear. Clarify
  before automating.
Presets

Five QA prompt structures to explore.

Each preset shows how role, output format, and coverage toggles combine for a common QA task. See what each option does and which choices matter.

How it works

Three steps to understanding your options.

1

Pick a purpose

Choose the type of QA output you want — test cases, bug report, Playwright spec. The relevant selectors pre-load, showing you the key options that matter for that task.

2

Add your content (optional)

Optionally paste your Jira ticket, steps, or API spec to see the prompt with real input. Content is not stored. Leave it blank to explore the structure on its own — add your actual content before sending to your AI agent.

3

Send to your AI agent

Copy the assembled prompt into ChatGPT, Claude, or any AI agent. Review and adjust your content as needed — the prompt structure is your reference, the input is always yours to control.

Build awareness of your prompting options.

Explore the structure, selectors, and options that make a quality QA prompt. Free, no account needed.