How can I reduce the download size for quick local tests?

The skill allows you to apply a sample-ratio (like 0.01 for 1%) and provides an algorithm override to reduce the number of training days, making downloads up to 100x faster.

Does it support complex algorithms with multiple dependencies?

Yes, it is specifically designed to detect parent-child algorithm structures and will identify and download all required data for both parent and sub-components.

Does this skill modify my Hotvect configuration?

No, the skill follows a strict configuration protection policy; it reads ~/.hotvect/config.json to construct commands but never modifies your local configuration files.

What happens if my AWS credentials expire during a download?

The skill checks your AWS identity before starting and provides specific CLI commands to refresh your credentials if they are missing or expired.

Hotvect Data Dependency Downloader

Name: Hotvect Data Dependency Downloader
Author: hotvect

byhotvect

•

Data Science & ML

Automates the retrieval of training and test data dependencies from S3 for local machine learning model development and backtesting.

This Claude Code skill streamlines the process of preparing local environments for Hotvect-based machine learning tasks. By parsing algorithm definitions and configuration files, it automatically calculates required date ranges, handles AWS credential verification, and executes optimized data downloads. It features intelligent sampling and date-range overrides to facilitate rapid local testing while ensuring data consistency between production and development environments, specifically optimized for Vowpal Wabbit workflows.

Key Features

01Integrated AWS credential verification and guided login workflows

02Support for data sampling (e.g., 1%) to enable fast local iteration and debugging

032 GitHub stars

04Automatic calculation of training and test date ranges based on lag and duration

05Configuration-driven command construction using local hotvect config files

06Advanced handling of parent-child algorithm structures for multi-stage dependencies

Use Cases

01Setting up local environments for algorithm training with specific historical datasets

02Downloading small subsets of production data for quick feature engineering cycles

03Preparing specific test datasets for local backtesting of updated ML models

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add hotvect/hotvect data-dependency-download

For use in Claude.ai and ChatGPT

Download Skill