Research-grade prompting assessment

Measure Your Real Prompting Ability

This is a controlled assessment, not a playground. Guide an AI agent through three realistic missions and receive a detailed score across six dimensions of prompting skill.

Estimated duration: 15-20 minutes

How It Works

01

Complete 3 Missions

Guide an AI agent through realistic tasks that increase in ambiguity. Each mission tests different facets of prompting skill.

02

Process Matters

You are scored on how you prompt, not just the final output. Framing, steering, verification, and error recovery all count.

03

Get Your Score

Receive detailed feedback across six scoring dimensions with actionable insights to improve your prompting technique.

Anti-cheat integrity

Behavioral telemetry ensures authentic results

Adaptive difficulty

Hidden friction mechanisms test real skill, not luck

Research-grade scoring

Multi-dimensional evaluation by calibrated AI judges