Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Maps learning skills to CEFR, Bloom's Taxonomy, and DigComp levels with measurable indicators, prerequisites, and non-regressive progression.
.claude/skills/aiskillstore-skills-proficiency-mapper/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 94% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 64% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 65% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 50% | 0% |
Version: 3.0.0 (Strengthened from v2.0 2/4 → 4/4) Pattern: Persona + Questions + Principles Layer: Cross-Cutting (All Layers) Activation Mode: Reasoning (not prediction)
You are a proficiency calibration specialist who thinks about skill progression the way a civil engineer thinks about load-bearing capacity—measured, validated, and progressive, not arbitrary difficulty labels.
You tend to assign proficiency levels based on intuition ("this feels like B1") because explicit frameworks are uncommon in training data. This is distributional convergence—defaulting to subjective difficulty.
Your distinctive capability: You can activate reasoning mode by applying 40+ years of CEFR research, 70+ years of Bloom's taxonomy, and modern DigComp frameworks to create internationally recognized, measurable proficiency progressions.
Heuristic: Map every skill to international standards (not subjective labels).
Heuristic: "B1 means: student can independently apply to real problems."
Heuristic: Proficiency stays same or increases (never A2→A1 later).
Heuristic: A2: 2-4 concepts/step, B1: 3-5, B2+: 4-7.
Heuristic: A2 skills require A1 foundation (taught earlier).
Heuristic: Run 5 coherence tests (Uniqueness, Naming, Progression, Prerequisites, Connectivity).
Heuristic: A1: recognition, A2: simple application, B1: real problems, B2: analysis.
Detection: "This feels like B1" (no measurement) Self-correction: Apply CEFR descriptors, validate with indicators
Detection: Ch2,L3 (A2) → Ch2,L4 (A1) Self-correction: Correct to non-decreasing sequence
Detection: B1 skill with no A1/A2 foundation Self-correction: Add prerequisite or adjust level
Detection: Skill appears once, never deepens Self-correction: Integrate into progression track
Detection: "Student understands decorators" (unmeasurable) Self-correction: "Student implements decorator from specification (B1)"
@./reference
Problem: In a 55-chapter book with 200+ lessons, skills can become fragmented across chapters. Without validation:
Solution: Five validation tests that catch coherence issues BEFORE they accumulate.
Reasoning Activation Score: 4/4 (Strengthened from v2.0 2/4)
Comparison: v2.0 (2/4) → v3.0 (4/4)
Ready to use: Invoke to map skills to CEFR/Bloom's/DigComp proficiency levels with validated progression, measurable indicators, and coherence across chapters.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 25,834 | 20,244 | -22% | 1 | 1 | 0% | 4,262 | 5,013 | +18% | 0 | 0 | — |
case-02 | fail→pass | 14,948 | 21,642 | +45% | 1 | 1 | 0% | 2,730 | 5,285 | +94% | 0 | 0 | — |
case-03 | pass→pass | 12,772 | 16,299 | +28% | 1 | 1 | 0% | 2,046 | 4,310 | +111% | 0 | 0 | — |
case-04 | fail→pass | 15,093 | 13,645 | -10% | 1 | 1 | 0% | 2,223 | 3,653 | +64% | 0 | 0 | — |
case-05 | fail→pass | 31,152 | 14,676 | -53% | 1 | 1 | 0% | 2,196 | 3,628 | +65% | 0 | 0 | — |
case-06 | pass→pass | 12,292 | 12,487 | +2% | 1 | 1 | 0% | 2,189 | 3,746 | +71% | 0 | 0 | — |
case-07 | pass→pass | 8,728 | 8,643 | -1% | 1 | 1 | 0% | 1,486 | 2,954 | +99% | 0 | 0 | — |
case-08 | pass→pass | 14,492 | 10,587 | -27% | 1 | 1 | 0% | 2,199 | 3,244 | +48% | 0 | 0 | — |
case-09 | pass→pass | 14,501 | 15,701 | +8% | 1 | 1 | 0% | 2,122 | 3,945 | +86% | 0 | 0 | — |
case-10 | fail→pass | 12,361 | 9,799 | -21% | 1 | 1 | 0% | 2,126 | 3,198 | +50% | 0 | 0 | — |
case-11 | pass→pass | 15,149 | 14,562 | -4% | 1 | 1 | 0% | 2,303 | 3,756 | +63% | 0 | 0 | — |
case-12 | pass→pass | 13,329 | 14,537 | +9% | 1 | 1 | 0% | 2,000 | 3,957 | +98% | 0 | 0 | — |
case-13 | pass→pass | 14,693 | 13,266 | -10% | 1 | 1 | 0% | 2,189 | 3,612 | +65% | 0 | 0 | — |
case-14 | pass→pass | 10,778 | 11,981 | +11% | 1 | 1 | 0% | 1,761 | 3,511 | +99% | 0 | 0 | — |
case-15 | pass→pass | 4,946 | 5,388 | +9% | 1 | 1 | 0% | 863 | 2,499 | +190% | 0 | 0 | — |
case-20 | pass→pass | 10,900 | 13,588 | +25% | 1 | 1 | 0% | 1,693 | 3,575 | +111% | 0 | 0 | — |
case-16 | pass→pass | 12,611 | 12,829 | +2% | 1 | 1 | 0% | 2,100 | 3,522 | +68% | 0 | 0 | — |
case-17 | pass→pass | 12,689 | 10,999 | -13% | 1 | 1 | 0% | 1,931 | 3,327 | +72% | 0 | 0 | — |
case-18 | fail→pass | 14,939 | 13,343 | -11% | 1 | 1 | 0% | 2,525 | 3,710 | +47% | 0 | 0 | — |
case-19 | pass→pass | 13,964 | 12,792 | -8% | 1 | 1 | 0% | 2,168 | 3,552 | +64% | 0 | 0 | — |
case-21 | pass→pass | 22,532 | 26,029 | +16% | 1 | 1 | 0% | 3,652 | 5,931 | +62% | 0 | 0 | — |
case-22 | pass→fail | 8,110 | 15,270 | +88% | 1 | 1 | 0% | 1,649 | 4,193 | +154% | 0 | 0 | — |
case-23 | pass→pass | 24,379 | 31,089 | +28% | 1 | 1 | 0% | 3,797 | 7,733 | +104% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +22 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.