skills.rest
epiblastai
epiblastaiOFFICIAL
Explore all skills by this author →
GitHub

write-ingestion-script

Write ingestion scripts for finalized Homeobox multimodal atlas datasets with loader contracts and schema validation.

7|Updated Feb 21, 2026
Search Tags:#ingestion#schema validation#anndata#zarr#data loaders#homeobox#multimodal atlas
One-click install
npx skills add https://github.com/epiblastai/homeobox --skill write-ingestion-script
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: write-ingestion-script
Source: https://github.com/epiblastai/homeobox/tree/main/packages/polycomb/skills/write-ingestion-script
Command: npx skills add https://github.com/epiblastai/homeobox --skill write-ingestion-script

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill removes the repetitive and error-prone work of writing ingestion scripts for finalized multimodal collections, ensuring raw dataset files are streamed into a Homeobox atlas with correct feature registration and row alignment.

Core Features & Use Cases

  • Loader Generation: Define per-feature-space loaders that convert raw DATA files into Homeobox readers and return the required metadata.
  • Reader Selection: Choose built-in readers for AnnData, COO triplets, or fragments, with guidance for implementing custom readers when necessary.
  • Validation and Reproducibility: Enforce finalized-table contracts, feature schema compatibility, row ordering, idempotent ingestion, and optional post-ingestion feature-oriented copies.
  • Use Case: Given a finalized collection containing gene expression and image feature datasets, create loaders for each feature space, ingest them into one atlas, and inspect the resulting ingestion report.

Quick Start

Use the write-ingestion-script skill to create an ingestion script for my finalized collection, selecting appropriate readers for each feature space and validating the resulting report.

Frequently Asked Questions about write-ingestion-script

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

FAQPage Schema
How do I write an ingestion script to add harmonized datasets to a multimodal atlas?▼

To write an ingestion script for a multimodal atlas, define per-feature-space loaders that convert raw data files into Homeobox readers. This ensures finalized datasets are streamed correctly with feature registration and row alignment.

How does schema validation work when ingesting sparse matrices and dense features?▼

Schema validation enforces finalized-table contracts and feature schema compatibility during ingestion. It ensures sparse matrices, dense features, and fragments are correctly registered and aligned before being added to the atlas.

Can I use custom readers for dataset-specific file formats during atlas ingestion?▼

You can implement custom readers for dataset-specific file formats when built-in readers are insufficient. The ingestion script supports optional custom reader requirements alongside built-in readers for AnnData, COO triplets, and fragments.

What is the best way to ensure idempotent execution and row-order alignment for data loaders?▼

To ensure idempotent execution and row-order alignment, apply loader contracts that validate feature registry compatibility and finalized-table requirements. This guarantees repeated ingestion runs produce consistent atlas data without duplication.

Do I need to finalize tables and harmonize schemas before writing an ingestion script?▼

You must finalize tables and harmonize schemas before writing an ingestion script. The script processes finalized, harmonized datasets and enforces finalized-table contracts to ensure data integrity during atlas loading.

Why does my ingestion script fail feature registry validation for multimodal datasets?▼

Ingestion scripts fail feature registry validation when raw datasets do not satisfy loader contracts or feature schema compatibility. Ensure row ordering and finalized-table requirements are met to resolve validation errors.

Related Skills

Explore More Software Engineering Skills→
baratadiegobaratadiego
2

ecological-data-foundation

Ingest, clean, and harmonize ecological occurrence data from GBIF, iNaturalist, OBIS, and IUCN.

Community
Advanced
epiblastaiepiblastai
7

prepare-package-for-resolution

Stages raw biomedical tables into Lance for schema resolution.

Official
Advanced
ya-wayya-way

bio-spatial-transcriptomics-spatial-data-io

Loads and unifies spatial transcriptomics data into a single AnnData object.

Community
Advanced
HKUDSHKUDS
30.4k

akshare

Aggregate AKShare financial data for Chinese, US, and HK instruments.

Official
Intermediate