General Track · Foundation
A public Rubric, a precise score
Five Tasks, 20 points each. Each criterion earns 0 to 5 points.
How the work is assessed
0: absent or wrong. 1: major gaps. 2: substantial gaps. 3: mostly correct, with gaps.
4: correct, with minor gaps. 5: fully correct and useful.
AI evaluates the final work. Passing the Exam requires 70/100. AI grading can contain errors.
The Level test is guidance, not a certificate. The General Exam requires a Course purchase.
Task 1 · 20 points
Write a useful brief
- Correct dates and order number · 0–5
- Clear apology and next step · 0–5
- No invented promise · 0–5
- Professional, concise email · 0–5
Task 2 · 20 points
Summarize without losing facts
- Three accurate decisions · 0–5
- Two actions with owners and dates · 0–5
- Rejected budget is not presented as approved · 0–5
- Readable, short summary · 0–5
Task 3 · 20 points
Check a claim
- Prefers the newer primary excerpt for current features · 0–5
- Explains date and source limitations · 0–5
- Does not invent retention or support guarantees · 0–5
- Names what to verify before buying · 0–5
Task 4 · 20 points
Analyze a small table
- Totals: 400 visits, 42 signups, EUR 248 · 0–5
- Rates: A 10%, B 15%, C 5% · 0–5
- Costs: A EUR 5, B EUR 6, C EUR 10 · 0–5
- Recommendation acknowledges small samples · 0–5
Task 5 · 20 points
Plan a safe multi-step task
- Clear ordered steps and useful output · 0–5
- Checks numbers and missing information · 0–5
- Uses approved, non-sensitive inputs · 0–5
- Human approval before any external action · 0–5