Skip to content
Back to skills

Datasets And Io

ASecurity

"Load Squidpy datasets and local spatial inputs with safe cache,

  • 247 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added September 9, 2026
dataapi

Works with

  • api

Security analysis

A100/100

Pro scans all 5 files and shows the line behind each finding

Scanned September 9, 2026

npx -y skills add VectorSpaceLab/AREX-Skill --skill datasets-and-io --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Datasets And Io?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Datasets And Io
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/vectorspacelab-datasets-and-io-arex-skill/badge)](https://www.skillsdirectory.com/skills/vectorspacelab-datasets-and-io-arex-skill)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: datasets-and-io
description: "Load Squidpy datasets and local spatial inputs with safe cache,
  layout, AnnData, and SpatialData validation."
disable-model-invocation: true
metadata:
  disco-role: operating
license: BSD 3-Clause
---

# Squidpy Datasets and IO

Use this sub-skill when a task starts with Squidpy sample data, local spatial omics outputs, saved `.h5ad` files, or `SpatialData` tables that must be checked before analysis, imaging, or plotting.

## Route Here For

- Choosing between `squidpy.datasets` download/cache helpers and local no-download readers.
- Loading local 10x Visium, Vizgen, or Nanostring/CosMx outputs with `sq.read.*`.
- Validating Squidpy-ready `AnnData` structure: `.obsm['spatial']`, `.uns['spatial']`, `library_id`, image metadata, and scalefactors.
- Checking `SpatialData` table availability and `table_key` before routing to graph or experimental imaging workflows.
- Diagnosing registry, cache, network/hash, `spatial/` layout, non-numeric coordinates, or image/scalefactor problems.

## Start Here

1. Read `references/data-loading.md` to pick a no-download local reader or a registry-backed dataset loader.
2. Check exact public signatures and required file conventions in `references/api-reference.md`.
3. Use `scripts/check_spatial_adata.py --help` to validate a saved `.h5ad` or an importable callable returning `AnnData`/`SpatialData`.
4. Use `references/troubleshooting.md` when loading succeeds but downstream workflows cannot find coordinates, `library_id`, images, scalefactors, or `table_key`.

## Safe Defaults

- Prefer `sq.read.visium`, `sq.read.vizgen`, `sq.read.nanostring`, `anndata.read_h5ad`, or `spatialdata.read_zarr` for user-owned local data; these do not consult Squidpy's dataset registry.
- Treat `sq.datasets.*` calls as potentially networked unless the needed files already exist in the chosen cache directory.
- Pass an explicit cache or input path instead of relying on global Scanpy settings.
- For Visium, use `load_images=False` only as a count/metadata fallback; it does not create `.obsm['spatial']`, images, or scalefactors.

## Boundaries

- Route spatial graph construction and statistics to `graph-analysis` after coordinates and table routing are clear.
- Route stable `ImageContainer`, image processing, segmentation, and image features to `image-analysis`.
- Route plot rendering and visual styling to `visualization`.
- Route experimental `SpatialData` tiling, tissue detection, QC, stitching, and stain workflows to `experimental-imaging`.

Files in this skill

  • SKILL.md2.5 KB
  • references/api-reference.md6.4 KB
  • references/data-loading.md8.3 KB
  • references/troubleshooting.md7.6 KB
  • scripts/check_spatial_adata.py12.7 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.