Can I use this for cloud-specific platforms like AWS or GCP?

The skill includes specific integration points for managed services like AWS SageMaker, Google Vertex AI, and Azure ML, as well as cloud-agnostic Kubernetes-based serving.

Which orchestration tools does this skill support?

This skill provides guidance and implementation patterns for leading DAG-based orchestrators including Apache Airflow, Dagster, Kubeflow Pipelines, and Prefect.

What kind of deployment strategies are covered?

It provides templates and logic for sophisticated deployment methods including canary releases, blue-green deployments, shadow deployments, and automated rollback triggers.

Does it include model monitoring and drift detection?

Yes, it covers best practices for monitoring production traffic, detecting model performance regression, and implementing data drift alerts to trigger retraining.

Does this skill help with data versioning?

Yes, it emphasizes data lineage and versioning using tools like DVC or feature stores to ensure full reproducibility of every model training run.

ML Pipeline Workflow

Name: ML Pipeline Workflow
Author: kivilaid

bykivilaid

•

数据科学与机器学习

Orchestrates end-to-end MLOps pipelines from data preparation and model training to automated production deployment and monitoring.

The ML Pipeline Workflow skill provides a comprehensive framework for building production-grade machine learning pipelines with an emphasis on reproducibility and modularity. It guides users through the entire ML lifecycle, including DAG-based orchestration for tools like Airflow or Kubeflow, automated data validation, experiment tracking, and sophisticated deployment patterns such as canary or blue-green releases. By implementing these standardized patterns, developers can ensure their ML systems are observable, testable, and ready for high-scale production environments while maintaining strict data lineage and quality controls.

主要功能

01End-to-end DAG orchestration for Airflow, Dagster, and Kubeflow

02Integration patterns for experiment tracking and model registries

03Advanced deployment strategies including canary and blue-green rollouts

04Automated data validation, quality checks, and versioning

05Robust model validation frameworks and A/B testing infrastructure

066 GitHub stars

使用场景

01Building automated, reproducible training workflows from scratch

02Designing modular MLOps architectures for cloud-native environments

03Implementing continuous training pipelines triggered by data drift

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add kivilaid/plugin-marketplace ml-pipeline-workflow

For use in Claude.ai and ChatGPT

Download Skill