A/B Test Tracker

General Productivityprompt-version-controltask_descriptionversion_aversion_b

Including adversarial test cases and a decision framework prevents the common mistake of picking a winner based only on happy-path performance.

Prompt
I'm A/B testing two versions of a prompt for {{task_description}}.\n\nVersion A:\n{{version_a}}\n\nVersion B:\n{{version_b}}\n\nEvaluation criteria:\n{{evaluation_criteria}}\n\nDesign an A/B test plan that includes:\n1. Sample size recommendation (number of test cases)\n2. Test cases that cover normal, edge, and adversarial inputs\n3. Scoring rubric for each evaluation criterion (1-5 scale with descriptions)\n4. Statistical method for determining a winner\n5. A results template I can fill in as I run tests\n6. Decision framework: when to pick A, pick B, or iterate further

Variables to customize

{{task_description}}{{version_a}}{{version_b}}{{evaluation_criteria}}

Why this prompt works

Including adversarial test cases and a decision framework prevents the common mistake of picking a winner based only on happy-path performance.

What you get when you save this prompt

Your workspace unlocks powerful tools to iterate and improve.

AI OPTIMIZE

AI Optimization

One-click improvement with structure analysis and pattern suggestions.

VERSION DIFF

Version History

Track every edit. Compare versions side-by-side with word-level diffs.

ORGANIZE
Development
Code Review
Testing
Marketing

Folders & Tags

Organize your library with nested folders, tags, and drag-and-drop.

MCP
$ npm i -g @promptingbox/mcp
Claude · Cursor · ChatGPT

Use Everywhere

Access prompts from Claude, Cursor, ChatGPT & more via MCP integration.

Your prompts, organized

Save, version, and access your best prompts across ChatGPT, Claude, Cursor, and more.