GPTomics' bioSkills includes 17 MIT-licensed single-cell skills that cover the whole analysis: preprocessing, doublet detection, clustering, annotation, batch integration, trajectories and cell-cell communication. They work across Scanpy, Seurat and SingleCellExperiment, and they tell the agent which method fits the data, such as Harmony versus scVI or CellTypist versus SingleR.
All 17 install with one command from Cuttlefish.MD, HackerNoon's license-checked skills marketplace. They follow the open Agent Skills standard, so they run in Claude Code, Codex CLI and 30+ other tools, and each is indexed in HackerNoon's SKILL.md Library.
The 17 single-cell skills
| Skill | What it does |
|---|---|
| Integrates samples or batches with Harmony, scVI/scANVI, Seurat CCA/RPCA, fastMNN, Scanorama or BBKNN | |
| Annotates cell types from references with CellTypist, SingleR, Azimuth or scANVI | |
| Infers ligand-receptor communication with LIANA, plus CellPhoneDB | |
| Runs dimensionality reduction and graph-based clustering in Scanpy or Seurat | |
| Infers copy-number changes to separate malignant from normal cells | |
| Reads, writes and converts AnnData, Seurat and SingleCellExperiment objects | |
| Tests whether cell-type composition changed between conditions, using Milo and more | |
| Finds and removes doublets with scDblFinder or Scrublet | |
| Assigns cells to samples from CITE-seq HTOs or MULTI-seq tags | |
| Rebuilds lineage trees and clones from CRISPR scars and barcodes | |
| Finds cluster marker genes and assigns manual cell-type labels | |
| Infers metabolite-mediated cell-cell communication with MEBOCOST | |
| Integrates CITE-seq RNA+protein and Multiome RNA+ATAC data | |
| Analyzes Perturb-seq and CROP-seq single-cell CRISPR screens | |
| Handles QC, ambient RNA, normalization and feature selection | |
| Analyzes single-cell ATAC-seq with Signac, ArchR or SnapATAC2 | |
| Infers trajectories, pseudotime, RNA velocity and fate probabilities |
For the full run from Cell Ranger output to annotated cell types, the scrnaseq-pipeline workflow skill chains these together.
Install them
All 17 come with the Data & Science plugin, along with the rest of the 562 bioSkills:
/plugin marketplace add hackernoon/cuttlefish-agent-skills /plugin install data-science@cuttlefish For Codex CLI or another agent that supports the standard, clone the repo and copy plugins/data-science/skills/ into your agent's skills folder, such as .agents/skills/.
Each skill keeps its MIT license and an ATTRIBUTION.md crediting GPTomics and the source commit. Browse the full library at hackernoon.com/skills, and star Cuttlefish.MD on GitHub.
