Which ML libraries does this skill support?

It provides built-in support for popular gradient boosting frameworks including XGBoost, LightGBM, and CatBoost.

How does the retraining scheduler work?

The scheduler can trigger a retrain based on three factors: elapsed time (e.g., daily), a drop in model accuracy below a specific threshold, or detected feature distribution drift.

Can I use custom technical indicators for features?

Yes, the FeatureEngineer class is designed to be extensible, allowing you to add price, volume, volatility, and custom technical indicators to the model's training set.

FreqAI is a self-adaptive machine learning framework originally built for FreqTrade that automates the lifecycle of model training, prediction, and retraining on new data.

What are the primary dependencies for this skill?

This skill requires freqtrade, xgboost, lightgbm, scikit-learn, pandas, and numpy to be installed in your environment.

FreqAI ML Integration

Name: FreqAI ML Integration
Author: LEDGERGHOST90

byLEDGERGHOST90

•

データサイエンスとML

Implements a self-adaptive machine learning retraining framework for automated trading signals and market regime detection.

This skill integrates the FreqAI framework into Claude Code, enabling the development of self-correcting machine learning models that automatically retrain on fresh market data. By leveraging rolling training windows and model drift detection, it prevents overfitting and adapts to shifting market regimes in real-time. It provides a standardized pipeline for feature engineering, multi-model support including XGBoost and LightGBM, and automated retraining schedules based on performance thresholds or temporal triggers, making it ideal for high-stakes time-series forecasting and algorithmic signal generation.

主な機能

012 GitHub stars

02Advanced feature engineering for OHLCV, volume, and volatility data

03Automated model retraining on daily, weekly, or monthly schedules

04Regime-aware training logic to adapt to changing market conditions

05Multi-model support for XGBoost, LightGBM, and CatBoost classifiers

06Real-time model drift detection and performance-based triggers

ユースケース

01Developing predictive signal generators with automated maintenance and monitoring

02Building self-adaptive algorithmic trading strategies that evolve with the market

03Detecting and correcting machine learning model drift in financial datasets

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ledgerghost90/sovereign_shadow_3 freqai-integration

For use in Claude.ai and ChatGPT

Download Skill