skills.rest
jasrajtulsi
jasrajtulsiCOMMUNITY
Explore all skills by this author →
GitHub

timesfm-forecasting

Forecast univariate time series with TimesFM zero-shot inference.

Updated Jul 1, 2026
Search Tags:#time series#forecasting#zero-shot#prediction intervals#timesfm#xreg#hardware check
One-click install
npx skills add https://github.com/jasrajtulsi/GRAD-SCOPE --skill timesfm-forecasting-jasrajtulsi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: timesfm-forecasting
Source: https://github.com/jasrajtulsi/GRAD-SCOPE/tree/main/.claude/skills/timesfm-forecasting
Command: npx skills add https://github.com/jasrajtulsi/GRAD-SCOPE --skill timesfm-forecasting-jasrajtulsi

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires numpy, pandas, torch, timesfm, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill removes the need to train a custom forecasting model when you need reliable univariate time series predictions from historical data.

Core Features & Use Cases

  • Zero-shot forecasting: Generate point forecasts and prediction intervals directly from a single series or batches of series.
  • Hardware-aware safety checks: Verify RAM, GPU/VRAM, disk space, and Python/package readiness before loading the model.
  • Practical time-series workflows: Forecast sales, sensors, energy, vitals, weather, and other univariate signals, then export results from CSV inputs or scripted runs.
  • Optional covariates and diagnostics: Use exogenous drivers with XReg, handle anomalies via quantile bands, and compare forecast quality across series.

Quick Start

Ask the assistant to run the TimesFM preflight check and then forecast your time series with point estimates and prediction intervals from the provided 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 without training a model?▼

Zero-shot time series forecasting uses pre-trained TimesFM checkpoints to predict univariate historical series directly, bypassing custom model training. You provide historical data, and the Skill outputs point forecasts and prediction intervals.

Can I include external variables when forecasting univariate time series data?▼

Yes, univariate time series forecasting supports optional exogenous regressors through XReg. This allows you to incorporate external drivers alongside your historical data to influence the resulting point forecasts and prediction intervals.

What hardware do I need to run TimesFM inference locally?▼

Running TimesFM inference requires sufficient RAM, GPU/VRAM, and disk space to load the model. The workflow includes preflight hardware checks to verify resource availability and Python package readiness before attempting to run predictions.

Does this forecasting approach work with batch CSV inputs?▼

Yes, batch time series forecasting processes multiple series directly from CSV inputs. It generates point estimates and prediction intervals for each series, allowing you to compare forecast quality across batches of historical data.

How are prediction intervals calculated for time series anomalies?▼

Prediction intervals are calculated using quantile bands during zero-shot inference. These bands provide confidence intervals around point forecasts, helping you identify anomalies and handle uncertainty across univariate series.

What types of historical data are supported for univariate forecasting?▼

Univariate forecasting supports historical series like sales, sensors, energy, vitals, and weather data. The model processes these single-variable sequences to generate future point estimates and quantile prediction intervals.

Related Skills

Explore More Data & Analytics Skills→
JosephWoodallJosephWoodall
1

timesfm-forecasting

Forecast univariate time series with quantile prediction intervals.

Community
Advanced
mkurmanmkurman
321

timesfm-forecasting

Forecast univariate time series with zero training using TimesFM.

Community
Advanced
recallnetrecallnet
1

timesfm-forecasting

Forecast univariate time series with TimesFM zero-shot forecasting.

Official
Advanced
shushuznshushuzn
4

timesfm-forecasting

Forecast univariate time series with quantile prediction intervals.

Community
Intermediate