Skip to content

The Assessment Flow

Before generating your skill, SkillThis runs a quick assessment of your input to determine how much methodology it contains. This takes a few seconds and decides whether to generate immediately or ask follow-up questions first.

When you submit your expertise description, the AI evaluates it on methodology specificity and assigns a score from 0 to 100.

Score RangeWhat It MeansWhat Happens
60-100Input has enough concrete methodologyProceeds directly to skill generation
0-59Input needs more detailTriggers extraction questions

The assessment looks for five things:

  1. Step-by-step processes - “First I do X, then Y”
  2. Named tools and frameworks - Specific software, methodologies, or techniques
  3. Clear input/output examples - What goes in, what comes out
  4. Quantifiable outcomes - Numbers, metrics, timelines
  5. Decision trees or conditional workflows - “If X, then Y; otherwise Z”
Input submitted
Assessment (score 0-100)
├── Score ≥ 60 ──→ Generate skill directly
└── Score < 60 ──→ Extraction questions (3 questions)
├── Answer all 3 ──→ Generate with enriched context
└── Skip ──→ Generate with original input

When the assessment runs, you’ll see a brief loading state. If your input scores below 60, the assessment also returns:

  • Reasoning - A 1-2 sentence explanation of why your input needs more detail
  • Missing elements - Specific suggestions for what would improve your input

This reasoning is displayed at the top of the extraction questions screen so you understand what’s needed.

The assessment and generation use different AI models for different purposes:

StepPurposeSpeed
AssessmentQuick check of input qualityFast (1-2 seconds)
GenerationCreate and grade the full skillSlower (10-20 seconds)

The assessment is intentionally fast so you get feedback quickly. If your input is good enough, you skip straight to generation with no delay.