schema-harmonization

Harmonize LanceDB tables to a target Homeobox schema with audited CurationApplicator transactions.

7|Updated Feb 21, 2026
One-click install
npx skills add https://github.com/epiblastai/homeobox --skill schema-harmonization
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: schema-harmonization
Source: https://github.com/epiblastai/homeobox/tree/main/packages/polycomb/skills/schema-harmonization
Command: npx skills add https://github.com/epiblastai/homeobox --skill schema-harmonization

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires lancedb, pandas, homeobox, polycomb, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This skill aligns raw collection- and dataset-level LanceDB tables with a target Homeobox schema while resolving inconsistent identifiers, labels, and metadata into canonical values without sacrificing provenance or auditability.

Core Features & Use Cases

  • Schema Alignment: Rename, cast, add, merge, and remove columns so raw tables conform to Homeobox schema classes.
  • Canonical Resolution: Resolve genes, proteins, molecules, ontology terms, organisms, perturbations, guides, and other biological identifiers through registered or custom resolution workflows.
  • Audited Transactions: Apply every harmonization change through CurationApplicator with dry-run validation, column guardrails, provenance, and SQLite audit records.
  • Advanced Workflows: Handle multi-field resolver fan-out, nullable-field recovery, publication and dataset metadata, genetic perturbation reshaping, and organism-specific ontology resolution.
  • Use Case: Harmonize heterogeneous single-cell datasets by aligning their metadata columns, canonicalizing cell types and organisms, resolving gene identifiers, and recording every mutation for reproducible downstream ingestion.

Quick Start

Ask the skill to dry-run a schema harmonization plan for a specified LanceDB table using the target Homeobox schema, resolving applicable ontology and cross-reference fields while reporting unresolved values and proposed audited changes.

Frequently Asked Questions about schema-harmonization

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I harmonize heterogeneous biomedical data schemas for single-cell atlases?

Schema harmonization aligns raw LanceDB tables with a target Homeobox schema by renaming, casting, and merging columns so heterogeneous single-cell atlas metadata conforms to canonical classes without losing provenance.

What is the best way to resolve biological identifiers and ontology terms during data curation?

Biological identifier resolution maps genes, proteins, molecules, and ontology terms to canonical values through registered or custom resolver workflows, handling multi-field fan-out and organism-specific ontology resolution during data curation.

Can I dry-run schema harmonization changes before applying them to LanceDB tables?

Yes, you can dry-run schema harmonization plans to validate proposed column changes and ontology resolutions against LanceDB tables, reporting unresolved values and audited mutations before committing to the dataset.

How do I maintain auditability and provenance when modifying biomedical atlas metadata?

Auditable biomedical metadata modification requires applying all harmonization changes through CurationApplicator transactions, which enforce column guardrails, validate schemas, and record provenance in SQLite audit logs for reproducible downstream ingestion.

Does schema harmonization support genetic perturbation data and nullable field recovery?

Yes, schema harmonization supports genetic perturbation reshaping and nullable-field recovery workflows, allowing complex biomedical atlas ingestion to handle missing metadata and restructure perturbation, guide, and publication data correctly.