Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Get current weather and forecasts (no API key required).
.claude/skills/kxiandaoyan-weather/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | -24% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -25% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 66% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 21% | 0% |
No environment variables or API keys required. This skill uses free public APIs:
curl command must be available in the systemTwo free services, no API keys needed.
Quick one-liner:
bashcurl -s "wttr.in/London?format=3" # Output: London: ⛅️ +8°C
Compact format:
bashcurl -s "wttr.in/London?format=%l:+%c+%t+%h+%w" # Output: London: ⛅️ +8°C 71% ↙5km/h
Full forecast:
bashcurl -s "wttr.in/London?T"
Format codes: %c condition · %t temp · %h humidity · %w wind · %l location · %m moon
Tips:
wttr.in/New+Yorkwttr.in/JFK?m (metric) ?u (USCS)?1 · Current only: ?0curl -s "wttr.in/Berlin.png" -o /tmp/weather.pngFree, no key, good for programmatic use:
bashcurl -s "https://api.open-meteo.com/v1/forecast?latitude=51.5&longitude=-0.12¤t_weather=true"
Find coordinates for a city, then query. Returns JSON with temp, windspeed, weathercode.
Docs: https://open-meteo.com/en/docs
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 4,912 | 3,375 | -31% | 1 | 1 | 0% | 768 | 912 | +19% | 0 | 0 | — |
case-02 | pass→pass | 5,305 | 2,484 | -53% | 1 | 1 | 0% | 917 | 701 | -24% | 0 | 0 | — |
case-03 | pass→pass | 5,541 | 1,881 | -66% | 1 | 1 | 0% | 895 | 675 | -25% | 0 | 0 | — |
case-04 | pass→pass | 3,083 | 2,783 | -10% | 1 | 1 | 0% | 495 | 820 | +66% | 0 | 0 | — |
case-05 | pass→pass | 3,494 | 2,067 | -41% | 1 | 1 | 0% | 566 | 687 | +21% | 0 | 0 | — |
case-06 | pass→pass | 6,301 | 3,101 | -51% | 1 | 1 | 0% | 1,026 | 883 | -14% | 0 | 0 | — |
case-07 | pass→pass | 4,740 | 2,184 | -54% | 1 | 1 | 0% | 639 | 731 | +14% | 0 | 0 | — |
case-08 | pass→pass | 4,408 | 1,826 | -59% | 1 | 1 | 0% | 733 | 709 | -3% | 0 | 0 | — |
case-09 | pass→pass | 3,092 | 2,300 | -26% | 1 | 1 | 0% | 506 | 770 | +52% | 0 | 0 | — |
case-10 | pass→pass | 3,849 | 2,148 | -44% | 1 | 1 | 0% | 600 | 735 | +23% | 0 | 0 | — |
case-11 | pass→pass | 5,004 | 2,093 | -58% | 1 | 1 | 0% | 841 | 714 | -15% | 0 | 0 | — |
case-12 | pass→pass | 4,141 | 7,422 | +79% | 1 | 1 | 0% | 626 | 1,097 | +75% | 0 | 0 | — |
case-13 | pass→pass | 3,988 | 3,914 | -2% | 1 | 1 | 0% | 687 | 1,065 | +55% | 0 | 0 | — |
case-14 | pass→pass | 5,571 | 2,722 | -51% | 1 | 1 | 0% | 905 | 788 | -13% | 0 | 0 | — |
case-15 | pass→pass | 6,730 | 3,761 | -44% | 1 | 1 | 0% | 1,137 | 1,059 | -7% | 0 | 0 | — |
case-16 | pass→pass | 2,224 | 2,097 | -6% | 1 | 1 | 0% | 336 | 763 | +127% | 0 | 0 | — |
case-17 | pass→pass | 3,619 | 2,606 | -28% | 1 | 1 | 0% | 648 | 790 | +22% | 0 | 0 | — |
case-18 | pass→pass | 2,296 | 1,965 | -14% | 1 | 1 | 0% | 351 | 690 | +97% | 0 | 0 | — |
case-19 | pass→pass | 5,074 | 4,993 | -2% | 1 | 1 | 0% | 839 | 1,294 | +54% | 0 | 0 | — |
case-20 | pass→pass | 15,087 | 8,189 | -46% | 1 | 1 | 0% | 2,529 | 1,828 | -28% | 0 | 0 | — |
case-21 | fail→fail | 14,899 | 8,550 | -43% | 1 | 1 | 0% | 2,381 | 1,847 | -22% | 0 | 0 | — |
case-22 | pass→pass | 8,859 | 7,673 | -13% | 1 | 1 | 0% | 1,564 | 1,703 | +9% | 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 +5 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.