skills.rest
sagunkayastha
sagunkayasthaCOMMUNITY
Explore all skills by this author →
GitHub

timesfm-forecasting

Forecast univariate time series with Google's TimesFM foundation model.

Updated Mar 15, 2026
Search Tags:#forecasting#zero-shot#prediction intervals#timeseries#foundation model#timesfm
One-click install
npx skills add https://github.com/sagunkayastha/claude_skills_collection --skill timesfm-forecasting-sagunkayastha
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: timesfm-forecasting
Source: https://github.com/sagunkayastha/claude_skills_collection/tree/main/machine-learning-ai/timesfm-forecasting
Command: npx skills add https://github.com/sagunkayastha/claude_skills_collection --skill timesfm-forecasting-sagunkayastha

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires timesfm, torch, numpy, pandas, matplotlib, scikit-learn, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill automates time series forecasting using Google's TimesFM foundation model, eliminating the need for custom model training for univariate time series data.

Core Features & Use Cases

  • Zero-Shot Forecasting: Predict future values for sales, sensor data, energy usage, weather, etc., without training.
  • Probabilistic Forecasts: Generates point forecasts and calibrated prediction intervals.
  • Versatile Input: Supports CSV, DataFrame, and array inputs.
  • System Checker: Includes a preflight script to ensure sufficient RAM/GPU resources before model loading.
  • Use Case: Forecast next 12 months of weekly sales data from a CSV file automatically.

Quick Start

Use the timesfm-forecasting skill to forecast the next 24 steps for the provided time series data.

Frequently Asked Questions about timesfm-forecasting

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate zero-shot time series forecasts for univariate data?▼

To generate zero-shot time series forecasts, this Skill uses Google's TimesFM foundation model to predict future values for univariate data without requiring custom model training.

Can I forecast directly from a CSV file without training a model first?▼

Yes, you can forecast directly from CSV, DataFrame, or array inputs. The Skill processes these formats to generate point forecasts and calibrated prediction intervals automatically.

How do prediction intervals work in TimesFM forecasting?▼

Prediction intervals in TimesFM forecasting provide probabilistic forecasts, generating both point estimates and calibrated intervals to quantify uncertainty in your time series predictions.

What are the system requirements for loading the TimesFM foundation model?▼

Loading the TimesFM foundation model requires sufficient RAM and GPU resources. A preflight system checker script verifies your environment meets these specifications before model loading.

Does zero-shot forecasting work for predicting weekly sales data?▼

Yes, zero-shot forecasting works for predicting weekly sales data. It automatically forecasts future values, such as the next 12 months of sales, without needing historical model training.

What are the limitations of using a foundation model for time series prediction?▼

A limitation of this foundation model approach is that it focuses exclusively on univariate time series data, meaning it does not natively handle multivariate inputs or covariates.

Related Skills

Explore More Data & Analytics Skills→
AlterLab-IEUAlterLab-IEU
58

alterlab-timesfm

Forecast univariate time-series data with zero-shot TimesFM point and quantile forecasts.

Community
Advanced
vasilyu1983vasilyu1983
73

ai-ml-timeseries

Forecast time series data with LightGBM, Transformers, and SHAP.

Community
Advanced
andikarachmanandikarachman
14

data-preprocessing

Preprocess raw data into clean, analysis-ready datasets with pandas and numpy.

Community
Advanced
yf8578yf8578
1

timesfm-forecasting

...

Community
Advanced