Make AI readiness measurable.
The Dark Factory Score
The Dark Factory Score is a 0-5 scale measuring how effectively you work with AI. Each level represents a distinct capability threshold, from manual coding to fully autonomous AI system design.
| Level | Name | Description |
|---|---|---|
| 0 | Manual | No AI proficiency. Writes all code by hand, doesn't use AI tools. |
| 1 | Autocomplete User | Uses basic line completions (Copilot tab-complete). Accepts or rejects suggestions without much evaluation. |
| 2 | Pair Programmer | Effective with chat-based AI coding. Can write decent prompts, evaluate AI-generated code, and integrate it into projects. Knows when AI output is wrong. |
| 3 | Agent Director | Directs multi-step AI agents effectively. Writes clear specs and task breakdowns. Reviews agent output critically. Understands tool use and context management. |
| 4 | Spec Author | Writes specs that AI executes autonomously into deployed features. Designs AI-integrated workflows. Manages context engineering across complex projects. |
| 5 | Dark Factory Architect | Designs fully autonomous AI development systems. Nobody's here yet -- but the assessment shows who's closest and what's left to close the gap. |
Psychometric-Grade Assessment
AdaptiveTest Skills uses Item Response Theory (IRT), the same psychometric model behind the SAT and GRE. Our 3-Parameter Logistic (3PL) model accounts for question discrimination, difficulty, and guessing probability.
Questions adapt in real-time based on your responses. This means fewer questions achieve the same accuracy as longer fixed-form tests. Your ability estimate converges as you answer, giving you a precise score in less time.
Our scoring is transparent. Each of the six skill domains is measured independently, and your overall Dark Factory Score is the weighted average across assessed domains.
Same Scale, Complementary Products
The Dark Factory Score uses the same 0-5 scale as the Dark Agent Factory Directory, where AI tools are rated on the same scale. Rate the tools and the people using them. Your assessment results include personalized tool recommendations from the DAF Directory matched to your skill level.
Built by Woodstock Software
AdaptiveTest Skills is built by Woodstock Software LLC, the same team behind the Dark Agent Factory Directory. We build tools that make AI adoption measurable and actionable for engineering teams.