Self-hosted alternatives to ml-pipeline

M

ml-pipeline

Designs and implements production-grade ML pipeline infrastructure: configures

ml-pipeline is an agent skill from the jeffallan claude-skills collection for designing and implementing production machine learning pipeline infrastructure. Its listed coverage includes experiment tracking with MLflow or Weights and Biases, orchestration with Kubeflow or Airflow, feature engineering, distributed training and hyperparameter tuning. Its persona is a senior ML pipeline engineer. Its workflow runs in six stages. The agent maps the data flow and defines the interfaces between components, then validates the data schema and its distribution before any training starts, and halts and reports if the checks fail. It implements feature engineering with transformation pipelines, feature stores and validation checks. It configures distributed training, hyperparameter tuning and resource allocation. It logs metrics, parameters and artefacts so experiments can be compared and reproduced. The last stage runs model evaluation gates and proposes A/B testing or shadow deployment before a model is promoted. The skill suits ML engineers and platform teams who need repeatable training and deployment rather than notebooks. It is listed on skills.sh, where the page shows install counts and GitHub stars for the collection. Its Gen Agent Trust Hub audit is marked as failing, which the FAQ lists, while Socket and Snyk pass. Install it with npx skills add https://github.com/jeffallan/claude-skills --skill ml-pipeline.