Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Show drill-me learning progress — topics studied, cards due for review, weakest concepts, and what to study next. Use when the user asks what's due, how their learning is going, or for their drill-me status.
.claude/skills/davepoon-drill-status/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -63% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -64% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -69% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -51% | 0% |
Report on the user's learning state. Read-only — no teaching, no ledger writes.
date +%Y-%m-%d.~/.drill-me/index.md and every file in ~/.drill-me/topics/.If the directory doesn't exist, say so and suggest /drill:me <topic> to start.
cw (confident-wrong) or leech, and any card withstability under 2 days after 3+ reviews — name the concepts plainly.
a 10-minute review today beats 30 minutes on Friday").
Keep it under a screenful. No lectures, no science explainers — just the dashboard.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,222 | 5,305 | +2% | 1 | 1 | 0% | 806 | 590 | -27% | 0 | 0 | — |
case-02 | fail→fail | 6,270 | 5,709 | -9% | 1 | 1 | 0% | 1,120 | 563 | -50% | 0 | 0 | — |
case-03 | fail→fail | 4,377 | 4,857 | +11% | 1 | 1 | 0% | 710 | 555 | -22% | 0 | 0 | — |
case-04 | fail→pass | 10,560 | 2,644 | -75% | 1 | 1 | 0% | 1,865 | 691 | -63% | 0 | 0 | — |
case-05 | pass→fail | 22,356 | 4,714 | -79% | 1 | 1 | 0% | 4,139 | 515 | -88% | 0 | 0 | — |
case-06 | pass→pass | 11,704 | 7,215 | -38% | 1 | 1 | 0% | 1,933 | 1,588 | -18% | 0 | 0 | — |
case-07 | fail→fail | 10,431 | 16,464 | +58% | 1 | 1 | 0% | 1,850 | 684 | -63% | 0 | 0 | — |
case-08 | pass→pass | 6,302 | 2,790 | -56% | 1 | 1 | 0% | 1,209 | 789 | -35% | 0 | 0 | — |
case-09 | fail→pass | 10,484 | 2,710 | -74% | 1 | 1 | 0% | 2,064 | 748 | -64% | 0 | 0 | — |
case-10 | pass→fail | 8,163 | 2,193 | -73% | 1 | 1 | 0% | 1,320 | 578 | -56% | 0 | 0 | — |
case-11 | fail→fail | 10,951 | 6,768 | -38% | 1 | 1 | 0% | 1,871 | 534 | -71% | 0 | 0 | — |
case-12 | fail→fail | 7,529 | 6,120 | -19% | 1 | 1 | 0% | 1,460 | 556 | -62% | 0 | 0 | — |
case-13 | pass→fail | 10,082 | 2,056 | -80% | 1 | 1 | 0% | 1,757 | 563 | -68% | 0 | 0 | — |
case-14 | fail→pass | 10,736 | 2,084 | -81% | 1 | 1 | 0% | 1,846 | 567 | -69% | 0 | 0 | — |
case-15 | pass→pass | 10,463 | 3,719 | -64% | 1 | 1 | 0% | 1,742 | 855 | -51% | 0 | 0 | — |
case-16 | fail→pass | 4,817 | 2,394 | -50% | 1 | 1 | 0% | 842 | 638 | -24% | 0 | 0 | — |
case-17 | pass→pass | 7,014 | 6,686 | -5% | 1 | 1 | 0% | 1,269 | 606 | -52% | 0 | 0 | — |
case-18 | fail→pass | 9,976 | 3,145 | -68% | 1 | 1 | 0% | 1,665 | 812 | -51% | 0 | 0 | — |
case-19 | fail→pass | 15,494 | 2,914 | -81% | 1 | 1 | 0% | 2,882 | 777 | -73% | 0 | 0 | — |
case-20 | pass→fail | 7,143 | 8,499 | +19% | 1 | 1 | 0% | 1,321 | 1,731 | +31% | 0 | 0 | — |
case-21 | pass→pass | 10,952 | 9,242 | -16% | 1 | 1 | 0% | 1,937 | 1,852 | -4% | 0 | 0 | — |
case-22 | pass→pass | 12,203 | 8,396 | -31% | 1 | 1 | 0% | 2,319 | 1,901 | -18% | 0 | 0 | — |
case-23 | pass→pass | 17,288 | 16,236 | -6% | 1 | 1 | 0% | 3,638 | 3,518 | -3% | 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 16 counted toward the lift figure. The other 7 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 +9 percentage points is the difference between those two pass rates over the 16 comparable cases. 4 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.