timesfm-forecasting

Performs zero-shot univariate time series forecasting with Google's TimesFM.

Updated Mar 10, 2026

Search Tags:#time series #forecasting #zero-shot #prediction intervals #foundation model #timesfm

Ready to run

npx skills add https://github.com/felixboehm/biochem-allergy --skill timesfm-forecasting-felixboehm

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: timesfm-forecasting
Source: https://github.com/felixboehm/biochem-allergy/tree/main/.claude/skills/timesfm-forecasting
Command: npx skills add https://github.com/felixboehm/biochem-allergy --skill timesfm-forecasting-felixboehm

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill automates time series forecasting for any univariate data without requiring custom model training, saving significant time and resources.

Core Features & Use Cases

Zero-Shot Forecasting: Predicts future values for sales, sensor data, energy usage, weather, and more, using a pre-trained foundation model.
Probabilistic Forecasts: Generates calibrated prediction intervals (quantiles) alongside point forecasts, providing a measure of uncertainty.
Batch Processing: Efficiently forecasts hundreds or thousands of time series simultaneously.
Use Case: Forecast next quarter's sales for 500 different product lines based on their historical weekly sales data, including 80% prediction intervals.

Quick Start

Use the timesfm-forecasting skill to forecast the next 24 data points 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 perform zero-shot time series forecasting without training a custom model?▼

Zero-shot time series forecasting uses a pre-trained foundation model to predict future values without custom training. This Skill uses Google's TimesFM to generate forecasts directly from your historical data, saving significant time and resources.

Can I forecast multiple time series in batch from a CSV or DataFrame?▼

Yes, you can forecast multiple time series in batch from CSV, DataFrame, or array inputs. The Skill efficiently processes hundreds or thousands of univariate time series simultaneously to generate point forecasts.

How do I get prediction intervals for my univariate time series forecasts?▼

Prediction intervals for univariate time series forecasting are generated as calibrated quantiles alongside point forecasts. This provides a measure of uncertainty for your predictions across domains like sales, sensors, energy, and weather.

What are the RAM or GPU requirements for running a foundation model for forecasting?▼

Running a foundation model for forecasting requires sufficient RAM or GPU resources. This Skill includes a preflight system checker that verifies your available RAM and GPU before model inference to ensure smooth execution.

What is the best way to forecast next quarter's sales for hundreds of product lines?▼

The best way to forecast sales for hundreds of product lines is batch processing with a zero-shot foundation model. You can input historical weekly sales data to generate point forecasts and 80% prediction intervals for all products simultaneously.