forecasting-time-series-data

This skill enables Claude to forecast future values based on historical time series data. It analyzes time-dependent data to identify trends, seasonality, and other patterns. Use this skill when the user asks to predict future values of a time series, analyze trends in data over time, or requires insights into time-dependent data. Trigger terms include "forecast," "predict," "time series analysis," "future values," and requests involving temporal data.

Install with Tessl CLI

npx tessl i github:jeremylongshore/claude-code-plugins-plus-skills --skill forecasting-time-series-data

What are skills?

1.01x

Quality

51%

Does it follow best practices?

Impact

96%

1.01x

Average score across 9 eval scenarios

Optimize this skill with Tessl

npx tessl skill review --optimize ./backups/skills-migration-20251108-070147/plugins/ai-ml/time-series-forecaster/skills/time-series-forecaster/SKILL.md

SKILL.md

Review

Evals

Evaluation results

75%

Retail Ice Cream Sales Forecast

Seasonal model selection and confidence intervals

Criteria

Without context

With context

Prophet selected

Seasonal motivation stated

20%

30%

Confidence intervals present

100%

Future values forecasted

100%

MAE reported

100%

RMSE reported

100%

Historical data used for training

100%

No arbitrary model override

60%

Forecast horizon specified

100%

Without context: $0.4129 · 1m 36s · 24 turns · 25 in / 5,084 out tokens

With context: $0.4835 · 1m 43s · 28 turns · 175 in / 5,603 out tokens

100%

Greenhouse Temperature Forecasting

Data preprocessing and characteristic analysis

Criteria

Without context

With context

Missing values handled

100%

Outlier treatment

100%

Trend identified

100%

Seasonality identified

100%

Autocorrelation examined

70%

100%

Model fit after preprocessing

100%

Confidence intervals included

100%

Evaluation metric reported

100%

Analysis summary produced

100%

Without context: $0.5008 · 2m 8s · 25 turns · 26 in / 7,408 out tokens

With context: $0.5402 · 2m 9s · 26 turns · 138 in / 7,220 out tokens

100%

Unemployment Rate Forecasting for Economic Planning

Stationary data forecasting and model evaluation

Criteria

Without context

With context

ARIMA model used

100%

Stationarity assessed

100%

Model order selection

100%

ACF or PACF used

30%

100%

Confidence intervals present

100%

MAE reported

100%

RMSE reported

100%

Forecast saved to file

100%

No Prophet as primary model

100%

Without context: $0.3720 · 2m 7s · 19 turns · 20 in / 5,943 out tokens

With context: $0.6347 · 2m 35s · 32 turns · 423 in / 8,390 out tokens

97%

-3%

Monthly Energy Consumption Forecast

Forecast visualization and integrated reporting

Criteria

Without context

With context

Forecast plot created

100%

Historical data in plot

100%

90%

Confidence band in plot

100%

90%

Confidence intervals in CSV

100%

MAE reported

100%

RMSE reported

100%

Model choice documented

100%

Seasonal model rationale

100%

Forecast saved to file

100%

Data quality step present

100%

85%

Without context: $0.5068 · 2m 44s · 24 turns · 26 in / 9,633 out tokens

With context: $0.5579 · 2m 12s · 30 turns · 63 in / 7,070 out tokens

100%

Weekly Store Revenue Forecasting

Model selection from data characteristics analysis

Criteria

Without context

With context

Trend identified

100%

Seasonality assessed

100%

Autocorrelation examined

100%

Stationarity assessed

100%

Model choice justified

100%

ARIMA selected for trend-dominant data

100%

Confidence intervals in forecast

100%

MAE reported

100%

RMSE reported

100%

Forecast covers 4 weeks

100%

Without context: $0.4168 · 2m 3s · 19 turns · 21 in / 7,097 out tokens

With context: $0.8488 · 3m 18s · 37 turns · 37 in / 10,759 out tokens

100%

Website Traffic Forecast for Content Planning

End-to-end forecast with data quality handling

Criteria

Without context

With context

Missing values handled

100%

Outlier addressed

100%

Model fit on cleaned data

100%

Weekly seasonality considered

100%

Model rationale documented

100%

30-day forecast produced

100%

Lower bound column present

100%

Upper bound column present

100%

MAE reported

100%

RMSE reported

100%

Without context: $0.4089 · 1m 56s · 18 turns · 18 in / 7,135 out tokens

With context: $0.6845 · 2m 29s · 31 turns · 63 in / 9,475 out tokens

100%

Quarterly Budget Revenue Forecast

Quarterly time series analysis and model selection

Criteria

Without context

With context

Trend identified

100%

Seasonality assessed

100%

Autocorrelation examined

100%

Model choice documented

100%

Model justified by data

100%

Non-seasonal model for trending data

100%

Confidence interval lower bound

100%

Confidence interval upper bound

100%

MAE reported

100%

RMSE reported

100%

Without context: $0.3593 · 1m 40s · 16 turns · 16 in / 6,687 out tokens

With context: $0.6475 · 2m 34s · 31 turns · 61 in / 9,964 out tokens

100%

Online Retail Hourly Order Volume Forecast

Multiple seasonality detection and Prophet selection

Criteria

Without context

With context

Weekly seasonality identified

100%

Annual/long-term pattern noted

100%

Multiple seasonality handled

100%

Model selection justified

100%

28-day forecast produced

100%

Lower bound present

100%

Upper bound present

100%

MAE reported

100%

RMSE reported

100%

Data analysis step present

100%

Without context: $0.7800 · 3m 27s · 25 turns · 25 in / 13,434 out tokens

With context: $0.8117 · 3m 18s · 30 turns · 209 in / 11,950 out tokens

100%

Manufacturing Defect Rate Reduction Forecast

Declining trend forecasting with ARIMA and evaluation

Criteria

Without context

With context

Downward trend identified

100%

Seasonality assessed as weak/absent

100%

Stationarity or autocorrelation examined

100%

Non-seasonal model selected

100%

Model selection justified by data

100%

6-month forecast produced

100%

Lower confidence bound present

100%

Upper confidence bound present

100%

MAE reported

100%

RMSE reported

100%

Without context: $0.4667 · 2m · 26 turns · 26 in / 6,944 out tokens

With context: $0.9013 · 3m 19s · 39 turns · 301 in / 11,802 out tokens

Evaluated: 5 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Retail Ice Cream Sales Forecast Greenhouse Temperature Forecasting Unemployment Rate Forecasting for Economic Planning Monthly Energy Consumption Forecast Weekly Store Revenue Forecasting Website Traffic Forecast for Content Planning Quarterly Budget Revenue Forecast Online Retail Hourly Order Volume Forecast Manufacturing Defect Rate Reduction Forecast

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.