ChatSpatial: Skill for Claude Code

.agents/skills/cell-composition/SKILL.md

cell-composition is a skill for Claude Code from cafferychen777/ChatSpatial. It costs 78 tokens per session (1,531 once invoked), scanned A, original, MIT.

An analysis of which cell types are present at each location in spatial data and the proportion of each type. For spot-based data, it estimates mixtures of cells; for single-cell data, it assigns cell types directly.

In plain words
What is it for?
Use it to estimate cell-type proportions in Visium or Slide-seq spots, annotate cells in Xenium, MERFISH, or CosMx data, and map cell distributions across tissue.
Why use it?
It helps interpret locations that contain several cells and shows where specific cell types are found. The method can be chosen based on the data's resolution and available reference information.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: agent in frontmatter; installed under .agents/ (shared by several agents).

This is cafferychen777/ChatSpatial's own configuration. It tells Claude Code how to work on ChatSpatial itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything ChatSpatial configures →

Reuse

Borrowing it

Nothing to install: this file belongs to cafferychen777/ChatSpatial. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/cafferychen777/ChatSpatial/main/.agents/skills/cell-composition/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/cafferychen777/ChatSpatial

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for cell-composition

README.md
[![agentmods](https://agentmods.dev/badge/skills/cafferychen777/chatspatial/cell-composition/github.svg)](https://agentmods.dev/skills/cafferychen777/chatspatial/cell-composition)
Your own site
<a href="https://agentmods.dev/skills/cafferychen777/chatspatial/cell-composition"><img src="https://agentmods.dev/badge/skills/cafferychen777/chatspatial/cell-composition/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for cell-composition

Your own site · 80×15
<a href="https://agentmods.dev/skills/cafferychen777/chatspatial/cell-composition"><img src="https://agentmods.dev/badge/skills/cafferychen777/chatspatial/cell-composition.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 78 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,531 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00078 $0.01531
Opus 5 $0.00039 $0.00766
Sonnet 5 $0.00016 $0.00306
Haiku 4.5 $0.00008 $0.00153

Measured 9d ago against content hash 430900ac1c23, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

cell-composition scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.agents/skills/cell-composition/SKILL.md · 187 lines

How it starts

The opening of the file, as written. The whole thing — 187 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Cell Composition Analysis

Overview

This skill answers the fundamental question: What cell types exist at each spatial location, and in what proportions?

For spot-based data (Visium), this requires deconvolution to estimate cell type mixtures. For single-cell resolution data (Xenium/MERFISH), this requires direct cell type annotation.

Decision Tree: Which Approach?

START: User wants cell type information
    │
    ├─ Q: What is the data resolution?
    │   │
    │   ├─ Spot-based (Visium, Slide-seq)
    │   │   └─ Q: Do you have a reference scRNA-seq dataset?
    │   │       ├─ YES → Deconvolution (see below)
    │   │       └─ NO → Q: Do you have marker genes?
    │   │           ├─ YES → Marker-based annotation
    │   │           └─ NO → Use public atlas as reference
    │   │
    │   └─ Single-cell (Xenium, MERFISH, CosMx)
    │       └─ Q: Do you have a reference dataset?
    │           ├─ YES → Transfer learning (Tangram/scANVI)
    │           └─ NO → Marker-based or LLM annotation
    │
    └─ Execute appropriate workflow

Deconvolution Method Selection

Quick Reference Table

Your Scenario Recommended Method Why
Quick exploration FlashDeconv Fastest, good accuracy
Publication quality RCTD (doublet mode) Gold standard, well-validated
Large dataset (>50k spots) Cell2location Scalable, GPU-accelerated
Need spatial imputation CARD Can impute cell-type-specific expression
No matched reference Tangram More flexible with reference
Deep learning preference DestVI/Stereoscope Variational inference

Detailed Method Guide

FlashDeconv (Recommended Default)
  • Speed: Fastest (~seconds to minutes)
  • Accuracy: Good for most applications
  • When to use: Initial exploration, iterative analysis
  • Limitations: Less accurate for rare cell types
RCTD (Publication Standard)
  • Speed: Moderate (~minutes)
  • Accuracy: Excellent, especially with doublet mode
  • When to use: Final results, manuscript figures
  • Modes:
    • doublet: High-resolution platforms (Visium HD, Slide-seq)
    • full: Standard Visium
    • multi: When spots may contain >2 cell types
  • Note: R-based, requires rpy2

Read the full file on GitHub · 187 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 9d ago First seen · 187 lines · 78 tokens per session scan A 430900ac1c23

Subscribe to this mod's changes

cell-composition is a skill published in the GitHub repository cafferychen777/ChatSpatial (44 stars, last pushed 24d ago), licensed MIT. It adds 78 tokens to every session and 1,531 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

cellxgene-census-query

Query CZ CELLxGENE Census (61M+ cells). Filter by cell type/tissue/disease, retrieve expression data, and integrate with scanpy/PyTorch for population-scale single-cell analysis. Use this skill when: (1) Querying single-cell expression data by cell type, tissue, or disease, (2) Exploring available single-cell datasets…

PharMolix/OpenBioMed · 105 tokens

single-cell-multi-omics-analysis-scvi

Probabilistic deep learning framework for single-cell multi-omics data analysis. Use this skill when: (1) Analyzing single-cell RNA-seq data with batch correction, (2) Integrating multi-modal data (CITE-seq, ATAC-seq, multi-omics), (3) Performing cell type annotation with scANVI, (4) Spatial transcriptomics…

PharMolix/OpenBioMed · 94 tokens

Disease Progression Trajectory Analysis

Use this skill when you have longitudinal patient omics data and want to.

TianGzlab/OmicsClaw · 6 tokens

Upstream Regulator Analysis

Identify transcription factors (TFs) driving observed differential expression by integrating ChIP-Atlas TF binding data (epigenomics) with RNA-seq DE results (transcriptomics). Ranks TFs by a combined regulatory score incorporating binding enrichment, target-DE overlap (Fisher's exact test), and directional…

TianGzlab/OmicsClaw · 5 tokens

bulkrna-coexpression

Load when discovering gene co-expression modules and hub genes in a bulk RNA-seq cohort via WGCNA-style soft-thresholded networks. Skip when direct DE comparison (use bulkrna-de); PPI lookup of an existing gene list (use bulkrna-ppi-network); single-cell co-expression (use sc-grn).

TianGzlab/OmicsClaw · 73 tokens

bulkrna-deconvolution

Load when estimating cell-type proportions in bulk RNA-seq samples from a single-cell or signature-matrix reference. Skip when the data is already single-cell (no deconvolution needed); spatial deconvolution (use spatial-deconv).

TianGzlab/OmicsClaw · 54 tokens