What is included in the generated notebook?

Each notebook includes data loading, exploratory analysis, model configuration, training, forecasting, metric evaluation (SMAPE, MASE), and professional visualizations.

Which Nixtla libraries are supported by this skill?

The skill supports StatsForecast for statistical models, MLForecast for machine learning approaches, and TimeGPT for foundation model forecasting via API.

Does this skill require a NIXTLA_API_KEY?

An API key is only required if you are generating demos specifically for TimeGPT. StatsForecast and MLForecast workflows can be generated and run using local open-source libraries.

Can I use my own datasets with the generator?

Yes, the skill allows you to specify custom datasets or use the built-in flags to generate sample data like the M4 competition series for testing.

Nixtla Demo Generator

Name: Nixtla Demo Generator
Author: intent-solutions-io

byintent-solutions-io

Data Science & ML

Generates production-ready Jupyter notebooks for time-series forecasting using Nixtla's StatsForecast, MLForecast, and TimeGPT libraries.

About

The Nixtla Demo Generator is a specialized Claude Code skill designed to streamline the creation of comprehensive time-series forecasting workflows. It automates the generation of interactive Jupyter notebooks that cover the entire data science pipeline—from data ingestion and preprocessing to model training, evaluation metrics, and professional visualizations. By supporting Nixtla's core libraries, this skill enables developers and data scientists to rapidly build proof-of-concepts, benchmarking notebooks, and educational content that follow production-grade best practices for statistical, machine learning, and foundation model forecasting.

Key Features

Automated generation of end-to-end forecasting pipelines in Jupyter
Built-in evaluation metrics including sMAPE, MASE, and MAE
Multi-library support for StatsForecast, MLForecast, and TimeGPT
Customizable data source integration and model configuration
0 GitHub stars
Production-ready templates with error handling and performance optimization

Use Cases

Creating rapid Proof of Concept (POC) demonstrations for stakeholders
Generating educational tutorials and documentation for forecasting workflows
Benchmarking different time-series models against standard datasets like M4

About

Key Features

Automated generation of end-to-end forecasting pipelines in Jupyter
Built-in evaluation metrics including sMAPE, MASE, and MAE
Multi-library support for StatsForecast, MLForecast, and TimeGPT
Customizable data source integration and model configuration
0 GitHub stars
Production-ready templates with error handling and performance optimization

Use Cases

Creating rapid Proof of Concept (POC) demonstrations for stakeholders
Generating educational tutorials and documentation for forecasting workflows
Benchmarking different time-series models against standard datasets like M4