Guide Metaxy feature definitions, versioning, and metadata tooling in data pipelines.
npx skills add https://github.com/anam-org/metaxy --skill metaxy
Please help me install this Agent Skill. Skill: metaxy Source: https://github.com/anam-org/metaxy/tree/main/.claude-plugin/skills/metaxy Command: npx skills add https://github.com/anam-org/metaxy --skill metaxy
Metaxy governs feature definitions, versioning, and metadata workflows for multi-modal data and ML pipelines. This skill provides practical guidance for creating feature definitions, configuring metadata stores, and using the Metaxy CLI to manage governance and lineage.
To begin, ask for a minimal example that defines a feature class using mx.BaseFeature and a simple FeatureSpec, including a basic id_columns and fields.
High-intent search queries and answers about installing and using this skill.
To define feature metadata for ML pipelines, you subclass mx.BaseFeature and configure a FeatureSpec with id_columns and fields. This structure allows you to declare field-level dependencies and track versions for reproducible data workflows.
The best way to configure a metadata store for feature versioning is by creating a metaxy.toml file. This configuration establishes reproducible pipelines by tracking metadata, governing feature definitions, and triggering downstream invalidations when fields change.
You manage feature lineage and migrations with the metaxy CLI by running its built-in commands to apply schema changes and validate feature graphs. This ensures metadata correctness and tracks dependencies across your data pipelines.
Yes, you can track field-level dependencies and downstream invalidations in data pipelines by defining them within your FeatureSpec. When a parent field version changes, the system uses these dependency graphs to trigger invalidations in downstream features.
To validate feature graphs and metadata, you should use testing patterns that verify dependency resolution and versioning correctness. The skill provides actionable templates to test BaseFeature subclasses and ensure your FeatureSpec configurations are accurate.
Yes, metaxy works for both small experiments and large multi-modal data pipelines. It provides scalable templates for feature definitions, dependency tracking, and data lineage governance, adapting to various pipeline sizes.
cdrguruInitialize Gemini CLI configuration and auditor persona for the PACK workflow.
KatlyngPlan and execute Convex schema migrations with the widen-narrow workflow.
AngelP17Automate end-to-end ML pipelines from data ingestion to deployment.
vivekganaDesign and govern data products with contracts and SLAs.
