Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Manage database schemas, Drizzle ORM, migrations, and data modeling. Use when creating tables, modifying columns, or planning database changes.
.claude/skills/aiskillstore-db-handler/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 56% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -47% | 0% |
pgTable definition in src/db/schema/{domain}.ts.relations and Foreign Keys.drizzle-kit generate).npx drizzle-kit push to sync schema changes directly to the database.userId, planId).WHERE clauses (e.g., status, email).createdAt).with: { ... }) or explicit .leftJoin() to fetch related data in a single query.export const roleEnum = ...).For detailed patterns, imports, and best practices, see reference.md.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 9,778 | 9,258 | -5% | 1 | 1 | 0% | 1,974 | 2,407 | +22% | 0 | 0 | — |
case-02 | fail→pass | 14,436 | 8,942 | -38% | 1 | 1 | 0% | 2,735 | 2,130 | -22% | 0 | 0 | — |
case-12 | fail→pass | 7,021 | 5,148 | -27% | 1 | 1 | 0% | 1,351 | 1,407 | +4% | 0 | 0 | — |
case-03 | fail→pass | 6,502 | 7,427 | +14% | 1 | 1 | 0% | 1,325 | 2,066 | +56% | 0 | 0 | — |
case-04 | fail→pass | 11,492 | 3,841 | -67% | 1 | 1 | 0% | 1,950 | 1,031 | -47% | 0 | 0 | — |
case-05 | fail→fail | 8,200 | 6,693 | -18% | 1 | 1 | 0% | 1,554 | 1,734 | +12% | 0 | 0 | — |
case-06 | fail→pass | 11,166 | 7,395 | -34% | 1 | 1 | 0% | 2,351 | 1,830 | -22% | 0 | 0 | — |
case-07 | fail→pass | 6,889 | 2,328 | -66% | 1 | 1 | 0% | 1,187 | 771 | -35% | 0 | 0 | — |
case-08 | pass→pass | 6,110 | 2,689 | -56% | 1 | 1 | 0% | 1,047 | 606 | -42% | 0 | 0 | — |
case-09 | pass→pass | 13,890 | 6,301 | -55% | 1 | 1 | 0% | 2,365 | 1,615 | -32% | 0 | 0 | — |
case-10 | pass→pass | 14,062 | 6,599 | -53% | 1 | 1 | 0% | 2,625 | 1,716 | -35% | 0 | 0 | — |
case-11 | pass→pass | 5,744 | 3,031 | -47% | 1 | 1 | 0% | 972 | 934 | -4% | 0 | 0 | — |
case-13 | pass→pass | 11,385 | 6,594 | -42% | 1 | 1 | 0% | 1,992 | 1,513 | -24% | 0 | 0 | — |
case-14 | pass→pass | 11,375 | 5,242 | -54% | 1 | 1 | 0% | 2,099 | 1,434 | -32% | 0 | 0 | — |
case-15 | fail→fail | 8,184 | 8,901 | +9% | 1 | 1 | 0% | 1,573 | 2,422 | +54% | 0 | 0 | — |
case-16 | pass→pass | 12,746 | 4,248 | -67% | 1 | 1 | 0% | 2,401 | 1,223 | -49% | 0 | 0 | — |
case-17 | pass→pass | 15,158 | 2,377 | -84% | 1 | 1 | 0% | 2,542 | 723 | -72% | 0 | 0 | — |
case-18 | fail→pass | 13,101 | 7,792 | -41% | 1 | 1 | 0% | 2,608 | 2,074 | -20% | 0 | 0 | — |
case-19 | fail→pass | 12,819 | 6,241 | -51% | 1 | 1 | 0% | 2,041 | 1,383 | -32% | 0 | 0 | — |
case-20 | pass→pass | 8,963 | 4,185 | -53% | 1 | 1 | 0% | 1,745 | 1,068 | -39% | 0 | 0 | — |
case-21 | pass→pass | 7,454 | 3,940 | -47% | 1 | 1 | 0% | 1,267 | 1,154 | -9% | 0 | 0 | — |
case-22 | pass→pass | 7,196 | 3,740 | -48% | 1 | 1 | 0% | 1,370 | 961 | -30% | 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. 22 cases were attempted. The headline lift of +41 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.