Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Meta Prophet — forecasting at scale. Additive model with yearly/weekly/daily seasonality, holiday effects, changepoints, and trend decomposition. Handles missing data and outliers automatically.
.claude/skills/mkurman-prophet/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✓→✓ | ＝ Same ✓ | -37% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -24% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 39% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | 1% | 0% |
| case-07 | ✓→✓ | ＝ Same ✓ | 22% | 0% |
Meta Prophet forecasts time series data with additive seasonality (yearly, weekly, daily), holiday effects, changepoint detection, and trend decomposition. Handles missing data and outliers automatically. Designed for business forecasting with human-interpretable components.
bashuv pip install prophet
pythonimport pandas as pd from prophet import Prophet import numpy as np df = pd.DataFrame({ "ds": pd.date_range("2023-01-01", periods=365, freq="D"), "y": [100 + i*0.5 + 10*(i%7==0) + np.random.normal(0, 5) for i in range(365)], }) model = Prophet(yearly_seasonality=True, weekly_seasonality=True) model.fit(df) future = model.make_future_dataframe(periods=90) forecast = model.predict(future) model.plot(forecast) model.plot_components(forecast)
pythonmodel = Prophet() model.add_country_holidays("US") model.fit(df)
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 7,200 | 7,437 | +3% | 1 | 1 | 0% | 1,359 | 1,360 | +0% | 0 | 0 | — |
case-02 | fail→fail | 13,131 | 8,273 | -37% | 1 | 1 | 0% | 2,231 | 1,710 | -23% | 0 | 0 | — |
case-03 | pass→pass | 5,008 | 2,133 | -57% | 1 | 1 | 0% | 772 | 484 | -37% | 0 | 0 | — |
case-04 | pass→pass | 9,379 | 5,405 | -42% | 1 | 1 | 0% | 1,569 | 1,188 | -24% | 0 | 0 | — |
case-05 | pass→pass | 3,439 | 2,816 | -18% | 1 | 1 | 0% | 550 | 765 | +39% | 0 | 0 | — |
case-06 | pass→pass | 4,239 | 2,540 | -40% | 1 | 1 | 0% | 681 | 689 | +1% | 0 | 0 | — |
case-07 | pass→pass | 3,864 | 2,522 | -35% | 1 | 1 | 0% | 544 | 665 | +22% | 0 | 0 | — |
case-08 | pass→pass | 3,678 | 2,968 | -19% | 1 | 1 | 0% | 376 | 667 | +77% | 0 | 0 | — |
case-09 | pass→pass | 5,846 | 3,770 | -36% | 1 | 1 | 0% | 1,079 | 936 | -13% | 0 | 0 | — |
case-10 | pass→pass | 6,602 | 5,154 | -22% | 1 | 1 | 0% | 1,305 | 1,380 | +6% | 0 | 0 | — |
case-11 | pass→pass | 3,202 | 3,063 | -4% | 1 | 1 | 0% | 492 | 751 | +53% | 0 | 0 | — |
case-12 | pass→pass | 5,972 | 3,158 | -47% | 1 | 1 | 0% | 1,011 | 802 | -21% | 0 | 0 | — |
case-13 | fail→fail | 4,368 | 1,433 | -67% | 1 | 1 | 0% | 628 | 468 | -25% | 0 | 0 | — |
case-14 | pass→pass | 3,498 | 3,192 | -9% | 1 | 1 | 0% | 547 | 784 | +43% | 0 | 0 | — |
case-15 | pass→pass | 2,920 | 2,274 | -22% | 1 | 1 | 0% | 429 | 648 | +51% | 0 | 0 | — |
case-16 | pass→pass | 5,479 | 1,907 | -65% | 1 | 1 | 0% | 1,000 | 634 | -37% | 0 | 0 | — |
case-17 | pass→pass | 7,476 | 5,103 | -32% | 1 | 1 | 0% | 1,224 | 1,061 | -13% | 0 | 0 | — |
case-18 | pass→pass | 6,444 | 3,776 | -41% | 1 | 1 | 0% | 1,078 | 907 | -16% | 0 | 0 | — |
case-19 | pass→pass | 2,885 | 2,553 | -12% | 1 | 1 | 0% | 487 | 660 | +36% | 0 | 0 | — |
case-20 | pass→pass | 6,208 | 4,684 | -25% | 1 | 1 | 0% | 1,132 | 1,129 | -0% | 0 | 0 | — |
case-21 | pass→pass | 13,746 | 8,361 | -39% | 1 | 1 | 0% | 2,424 | 1,893 | -22% | 0 | 0 | — |
case-22 | pass→pass | 10,436 | 6,759 | -35% | 1 | 1 | 0% | 1,977 | 1,440 | -27% | 0 | 0 | — |
case-23 | pass→pass | 11,758 | 10,192 | -13% | 1 | 1 | 0% | 2,268 | 2,367 | +4% | 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. The headline lift of 0 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Other measured skills in the registry, with their headline benchmark lift.