Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Report spaced-repetition statistics: total facts, overdue count, retention rate, active decks, and facts added this week. Use when the user says "stats", "estatísticas", "como está meu aprendizado", "/learn-stats", or wants a dashboard of their learning loop.
.claude/skills/evolution-foundation-learn-stats/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 55% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 105% | 0% |
Reads workspace/learning/facts/ and workspace/learning/.state/review-log.jsonl to produce a markdown stats report.
User wants a snapshot of their learning loop health.
.md files in workspace/learning/facts/total = count of all fact filesoverdue = count where next_review <= todayactive_decks = count of unique deck values across all factsadded_this_week = count where created >= today - 7 daysIf workspace/learning/facts/ does not exist or is empty: > "Nenhum fato encontrado. Use /learn-capture para adicionar fatos primeiro." > Stop here.
workspace/learning/.state/review-log.jsonl line by line (each line is a JSON object)total_reviews = 0, retention_rate = "N/A"total_reviews = count of all log entriesgood_easy_count = count where grade == "good" OR grade == "easy"retention_rate = round(good_easy_count / total_reviews * 100)%Print in markdown, in pt-BR:
markdown## 📊 Learning Loop — Stats | Métrica | Valor | |---------------------|--------------| | Total de fatos | {total} | | Vencidos hoje | {overdue} | | Taxa de retenção | {retention_rate} ({good_easy_count}/{total_reviews} revisões) | | Decks ativos | {active_decks} | | Adicionados (7 dias)| {added_this_week} | **Data:** {today YYYY-MM-DD}
If overdue > 0: > 💡 Use /learn-review para revisar os fatos vencidos.
If total < 10: > 💡 Use /learn-capture para adicionar mais fatos (meta v0: 20 fatos em 2 semanas).
If overdue > 0 and overdue <= 10, list them:
### Fatos vencidos
- `{filename}` (deck: {deck}, venceu: {next_review})
> {Retrieval Q — first 80 chars}N/A for retention — do not error.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 7,601 | 2,319 | -69% | 1 | 1 | 0% | 1,200 | 1,070 | -11% | 0 | 0 | — |
case-02 | fail→fail | 14,025 | 10,494 | -25% | 1 | 1 | 0% | 1,836 | 2,072 | +13% | 0 | 0 | — |
case-03 | fail→fail | 11,415 | 19,806 | +74% | 1 | 1 | 0% | 2,055 | 1,247 | -39% | 0 | 0 | — |
case-04 | fail→pass | 7,127 | 2,560 | -64% | 1 | 1 | 0% | 1,360 | 1,204 | -11% | 0 | 0 | — |
case-05 | fail→pass | 4,641 | 2,089 | -55% | 1 | 1 | 0% | 711 | 1,104 | +55% | 0 | 0 | — |
case-06 | fail→pass | 6,297 | 3,751 | -40% | 1 | 1 | 0% | 1,197 | 1,500 | +25% | 0 | 0 | — |
case-07 | pass→pass | 6,333 | 4,706 | -26% | 1 | 1 | 0% | 1,124 | 1,701 | +51% | 0 | 0 | — |
case-08 | pass→pass | 10,323 | 7,760 | -25% | 1 | 1 | 0% | 2,063 | 1,474 | -29% | 0 | 0 | — |
case-09 | fail→pass | 6,745 | 2,145 | -68% | 1 | 1 | 0% | 1,181 | 1,047 | -11% | 0 | 0 | — |
case-10 | fail→pass | 2,719 | 2,763 | +2% | 1 | 1 | 0% | 635 | 1,301 | +105% | 0 | 0 | — |
case-11 | pass→pass | 3,731 | 3,167 | -15% | 1 | 1 | 0% | 857 | 1,475 | +72% | 0 | 0 | — |
case-12 | pass→pass | 2,285 | 2,120 | -7% | 1 | 1 | 0% | 374 | 1,035 | +177% | 0 | 0 | — |
case-13 | fail→pass | 7,741 | 2,408 | -69% | 1 | 1 | 0% | 1,210 | 1,177 | -3% | 0 | 0 | — |
case-14 | fail→pass | 10,274 | 2,631 | -74% | 1 | 1 | 0% | 1,596 | 1,166 | -27% | 0 | 0 | — |
case-15 | fail→pass | 4,620 | 3,392 | -27% | 1 | 1 | 0% | 685 | 1,367 | +100% | 0 | 0 | — |
case-16 | fail→pass | 7,535 | 2,355 | -69% | 1 | 1 | 0% | 1,139 | 1,149 | +1% | 0 | 0 | — |
case-17 | fail→fail | 16,294 | 1,832 | -89% | 1 | 1 | 0% | 2,619 | 1,027 | -61% | 0 | 0 | — |
case-18 | pass→pass | 8,141 | 6,208 | -24% | 1 | 1 | 0% | 1,293 | 1,659 | +28% | 0 | 0 | — |
case-19 | pass→pass | 4,427 | 3,915 | -12% | 1 | 1 | 0% | 738 | 1,161 | +57% | 0 | 0 | — |
case-20 | pass→pass | 6,954 | 3,662 | -47% | 1 | 1 | 0% | 1,204 | 1,415 | +18% | 0 | 0 | — |
case-21 | fail→fail | 4,430 | 6,038 | +36% | 1 | 1 | 0% | 271 | 1,152 | +325% | 0 | 0 | — |
case-22 | fail→fail | 6,774 | 8,135 | +20% | 1 | 1 | 0% | 1,182 | 2,247 | +90% | 0 | 0 | — |
case-23 | fail→pass | 7,554 | 3,392 | -55% | 1 | 1 | 0% | 1,185 | 1,304 | +10% | 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, and 21 counted toward the lift figure. The other 2 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +43 percentage points is the difference between those two pass rates over the 21 comparable cases. 3 cases got worse with the skill loaded, and they are 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.