Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add HolobiomicsLab/asb-skill-collections --skill cluster-umap-layout-reproducibility-benchmarkinggit clone --depth 1 https://github.com/HolobiomicsLab/asb-skill-collectionsWrote 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.
[](https://agentmods.dev/skills/holobiomicslab/asb-skill-collections/cluster-umap-layout-reproducibility-benchmarking)<a href="https://agentmods.dev/skills/holobiomicslab/asb-skill-collections/cluster-umap-layout-reproducibility-benchmarking"><img src="https://agentmods.dev/badge/skills/holobiomicslab/asb-skill-collections/cluster-umap-layout-reproducibility-benchmarking/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.
<a href="https://agentmods.dev/skills/holobiomicslab/asb-skill-collections/cluster-umap-layout-reproducibility-benchmarking"><img src="https://agentmods.dev/badge/skills/holobiomicslab/asb-skill-collections/cluster-umap-layout-reproducibility-benchmarking.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00076 | $0.01860 |
| Opus 5 | $0.00038 | $0.00930 |
| Sonnet 5 | $0.00015 | $0.00372 |
| Haiku 4.5 | $0.00008 | $0.00186 |
Grade A, and why
cluster-umap-layout-reproducibility-benchmarking 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 12d 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.
How it starts
The opening of the file, as written. The whole thing — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
cluster-umap-layout-reproducibility-benchmarking
Summary
Verify that a single-cell ATAC-seq analysis pipeline produces reproducible cluster assignments and UMAP coordinates by comparing observed outputs against a documented reference implementation. This skill ensures that dimension reduction, clustering, and visualization steps have been correctly applied and yield consistent results.
When to use
Apply this skill when you have executed an end-to-end SnapATAC2 pipeline on the pbmc10k_multiome dataset (or a similar single-cell ATAC-seq dataset with a published reference) and need to validate that spectral embedding, Leiden clustering, and UMAP layout have converged to expected cluster identities and coordinate distributions. Use it after calling tl.spectral, tl.leiden, and tl.umap to confirm the pipeline has not diverged from the documented workflow.
When NOT to use
- Data has not yet been processed through pp.import_fragments and pp.add_tile_matrix—comparison requires fully preprocessed tile matrices.
- No reference implementation or published cluster assignments are available for the dataset—reproducibility benchmarking requires a ground truth to compare against.
- The dataset is substantially different from the reference (e.g., different cell type composition, different species, or very different sequencing depth)—reproducibility metrics may be uninformative if the underlying biology differs.
Inputs
- AnnData object with tile matrix generated via pp.add_tile_matrix (rows: cells, columns: genomic tiles; values: paired-insertion counts)
- Spectral eigenvectors computed by tl.spectral (shape: n_cells × n_components)
- Leiden cluster assignments (categorical, one per cell)
- UMAP coordinates (shape: n_cells × 2)
- Reference cluster assignments and UMAP layouts from published tutorial or prior validated run
Outputs
- Comparison metrics: adjusted Rand index (ARI), purity, or homogeneity between observed and reference cluster assignments
- Correlation coefficient (Pearson r) between observed and reference cluster-averaged UMAP coordinates
- Paired UMAP visualizations (observed vs. reference) for visual assessment of layout consistency
- Cluster composition table (number of cells per cluster, cell type annotations if available)
- Report indicating pass/fail status against reproducibility thresholds
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.
- 12d ago First seen · 114 lines · 76 tokens per session scan A 2ef06304119c
cluster-umap-layout-reproducibility-benchmarking is a skill published in the GitHub repository HolobiomicsLab/asb-skill-collections (15 stars, last pushed yesterday), licensed Apache-2.0. It adds 76 tokens to every session and 1,860 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.
Other skills, from other repositories
gsva-analysis-and-visualization
Use this skill to run GSVA or ssGSEA pathway-level differential analysis from a bulk expression matrix and a sample group file, then generate a heatmap from the saved GSVA result object. Trigger keywords: GSVA, ssGSEA, pathway enrichment, KEGG pathway analysis, MSigDB. NOT for: gene-level differential expression…
medical-research-literature-reader-pro
A medical-research-native literature reading skill for users with clinical, bioinformatics, translational, and basic experimental backgrounds. Use this skill whenever a user wants to read, analyze, critique, or interpret a medical or scientific paper — whether they provide a PDF, abstract, DOI, PMID, or just a title.…
anatomy-quiz-master
Generate interactive anatomy quizzes for medical education with multiple.
decision-curve-analysis
Use when evaluating the clinical utility of a binary prediction model from a single clinical CSV file by fitting a logistic decision-curve model, plotting decision and clinical-impact curves, and exporting summary outputs. NOT for: survival calibration, ROC-only discrimination analysis, nomogram construction, or…
elastic-net-feature-selection
Use when selecting predictive genes or other molecular features from bulk expression matrices for binary case-vs-control classification with elastic net logistic regression, including coefficient path and cross-validation plots. Trigger keywords: elastic net, glmnet, feature selection, binary classification…
external-model-validation
Use when validating an existing prognostic risk signature on an external bulk expression cohort with survival outcomes, producing risk scores, Kaplan-Meier curves, risk distribution plots, heatmap, and time-dependent ROC curves. NOT for: model training, feature selection, nomogram construction, calibration analysis…