Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use `uv` instead of pip/python/venv. Run scripts with `uv run script.py`, add deps with `uv add`, use inline script metadata for standalone scripts.
.claude/skills/dicklesworthstone-uv/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -60% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -58% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -69% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -58% | 0% |
bashuv run script.py # Run a script uv run --with requests script.py # Run with ad-hoc dependency uv add requests # Add dependency to project uv init --script foo.py # Create script with inline metadata
python# /// script # requires-python = ">=3.12" # dependencies = ["requests"] # ///
See scripts.md for full details on running scripts, locking, and reproducibility.
Use uv_build for pure Python packages:
toml[build-system] requires = ["uv_build>=0.9.28,<0.10.0"] build-backend = "uv_build"
See build.md for project structure, namespaces, and file inclusion.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 7,795 | 4,506 | -42% | 1 | 1 | 0% | 1,467 | 586 | -60% | 0 | 0 | — |
case-02 | fail→pass | 7,648 | 2,421 | -68% | 1 | 1 | 0% | 1,523 | 647 | -58% | 0 | 0 | — |
case-03 | pass→pass | 6,963 | 1,890 | -73% | 1 | 1 | 0% | 1,277 | 532 | -58% | 0 | 0 | — |
case-04 | pass→pass | 4,513 | 1,217 | -73% | 1 | 1 | 0% | 758 | 399 | -47% | 0 | 0 | — |
case-05 | pass→pass | 6,140 | 2,402 | -61% | 1 | 1 | 0% | 952 | 501 | -47% | 0 | 0 | — |
case-06 | pass→pass | 5,068 | 2,116 | -58% | 1 | 1 | 0% | 934 | 495 | -47% | 0 | 0 | — |
case-07 | pass→pass | 2,294 | 2,092 | -9% | 1 | 1 | 0% | 410 | 551 | +34% | 0 | 0 | — |
case-08 | pass→pass | 4,557 | 2,762 | -39% | 1 | 1 | 0% | 771 | 580 | -25% | 0 | 0 | — |
case-09 | pass→pass | 5,217 | 1,805 | -65% | 1 | 1 | 0% | 975 | 543 | -44% | 0 | 0 | — |
case-10 | pass→pass | 6,854 | 3,255 | -53% | 1 | 1 | 0% | 1,170 | 668 | -43% | 0 | 0 | — |
case-11 | fail→pass | 13,356 | 2,032 | -85% | 1 | 1 | 0% | 1,802 | 565 | -69% | 0 | 0 | — |
case-12 | pass→pass | 3,371 | 1,807 | -46% | 1 | 1 | 0% | 550 | 479 | -13% | 0 | 0 | — |
case-13 | pass→pass | 7,689 | 1,686 | -78% | 1 | 1 | 0% | 1,377 | 462 | -66% | 0 | 0 | — |
case-14 | fail→pass | 4,791 | 4,377 | -9% | 1 | 1 | 0% | 839 | 981 | +17% | 0 | 0 | — |
case-15 | pass→pass | 3,422 | 1,860 | -46% | 1 | 1 | 0% | 456 | 447 | -2% | 0 | 0 | — |
case-16 | pass→pass | 3,894 | 1,947 | -50% | 1 | 1 | 0% | 625 | 509 | -19% | 0 | 0 | — |
case-17 | pass→pass | 10,429 | 1,840 | -82% | 1 | 1 | 0% | 1,520 | 482 | -68% | 0 | 0 | — |
case-18 | pass→pass | 3,998 | 2,280 | -43% | 1 | 1 | 0% | 687 | 497 | -28% | 0 | 0 | — |
case-19 | fail→fail | 7,414 | 2,367 | -68% | 1 | 1 | 0% | 1,116 | 493 | -56% | 0 | 0 | — |
case-20 | pass→pass | 4,352 | 3,181 | -27% | 1 | 1 | 0% | 595 | 616 | +4% | 0 | 0 | — |
case-21 | pass→pass | 2,462 | 2,127 | -14% | 1 | 1 | 0% | 429 | 518 | +21% | 0 | 0 | — |
case-22 | pass→pass | 6,353 | 3,803 | -40% | 1 | 1 | 0% | 963 | 780 | -19% | 0 | 0 | — |
case-23 | pass→pass | 2,376 | 2,985 | +26% | 1 | 1 | 0% | 358 | 604 | +69% | 0 | 0 | — |
case-24 | pass→pass | 9,679 | 2,523 | -74% | 1 | 1 | 0% | 1,483 | 532 | -64% | 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. 24 cases were attempted. The headline lift of +17 percentage points is the difference between those two pass rates over the 24 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.