Business & Operations

Load when inferring TF → target gene regulatory networks on a normalised scRNA AnnData via pySCENIC (GRNBoost2 + cisTarget + AUCell) or correlation-based GRN fallback (when arboreto is unavailable, in --demo, or with --allow-simplified-grn). Skip when computing ligand-receptor cell-cell signalling (use sc-cell-communication) or for predicting genetic-KO effects (use sc-in-silico-perturbation).

Load when removing low-quality cells and lowly-detected genes from a single-cell AnnData using QC-derived thresholds or tissue presets. Skip for the full normalize→HVG→PCA→cluster pipeline (use sc-preprocessing) or when reads are still raw FASTQ (use sc-fastq-qc → sc-count first).

Load when checking raw single-cell FASTQ read quality (Phred / GC / adapter / length) before counting. Skip when reads are already counted (use sc-qc) or for bulk FASTQ (use bulkrna-read-qc).

Load when annotating putative doublets in single-cell RNA-seq using Scrublet, DoubletDetection, DoubletFinder, scDblFinder, or scds. Skip when ambient RNA is the contamination problem (use sc-ambient-removal) or before counts exist (use sc-fastq-qc / sc-count).

Load when turning scRNA FASTQ (or existing CellRanger/STARsolo/SimpleAF/kb-python output) into a downstream-ready AnnData. Skip when reads are already counted into AnnData (use sc-standardize-input) or for raw quality assessment only (use sc-fastq-qc).

Load when building the neighbour graph, embedding (UMAP/t-SNE/diffmap/PHATE), and clustering (Leiden/Louvain) on a normalised single-cell AnnData. Skip when QC/normalisation/HVG/PCA have not run yet (use sc-preprocessing) or for marker ranking after clustering (use sc-markers).

Load when computing cell-cell ligand-receptor communication on an annotated scRNA AnnData via builtin scorer, LIANA, CellPhoneDB, CellChat (R), or NicheNet (R). Skip when assigning cell-type labels (use sc-cell-annotation) or for transcription factor → target regulatory networks (use sc-grn).

Load when integrating multi-sample scRNA-seq with Harmony, scVI, scANVI, BBKNN, Scanorama, SIMBA, or supported R-backed methods to remove batch effects. Skip when the data is one sample (no batch effect to integrate) or for upstream merging only (use sc-multi-count).

Load when running an XCMS-style preprocessing summary on LC-MS metabolomics raw / vendor-converted files — emits a peak table with m/z, retention time, and per-sample intensities. Skip when working with an already-built peak table (use `metabolomics-peak-detection`) or when only annotation is needed (use `metabolomics-annotation`).

Load when summarising functional impact of an annotated variant CSV — per-IMPACT counts (HIGH / MODERATE / LOW / MODIFIER), top consequences, gene-affected count. Skip when input is a raw VCF (convert with `bcftools +split-vep` first), when calling raw variants (use `genomics-variant-calling`), or filtering VCFs (use `genomics-vcf-operations`).

Load when summarising a phased VCF (output of WhatsHap / SHAPEIT5 / Eagle2) — phased fraction of het variants, phase-block N50, PS-field parsing, pipe-delimited genotype detection. Skip when the input is unphased (run a phaser first) or when calling small variants (use `genomics-variant-calling`).

Load when summarising rMATS / SUPPA2 alternative-splicing output and identifying significant differential splicing events. Skip if you only have count-level DE (use bulkrna-de) or for splicing in single-cell or spatial data (currently unsupported).

Load when converting gene identifiers between Ensembl, Entrez, and HGNC symbol in a bulk RNA-seq count matrix. Skip if the input is already in the desired identifier system, for organisms outside human/mouse, or for non-bulk-counts inputs.

Load when running pathway / GO term enrichment on a bulk RNA-seq DE result list. Skip if the input is single-cell (use sc-enrichment), spatial (use spatial-enrichment), or for metabolite pathways (use metabolomics-pathway-enrichment).

Operate Google TxGemma prediction and chat models for therapeutic property prediction across small molecules, proteins, nucleic acids, diseases, targets, and cell lines. Use when formatting Therapeutics Data Commons tasks, choosing TxGemma model size or variant, running local or Model Garden inference, fine-tuning on private therapeutic data, or evaluating TxGemma in drug-discovery workflows.

Build and evaluate medical text and vision applications with Google MedGemma, including MedGemma 1.5 workflows for CT, MRI, whole-slide pathology, longitudinal chest X-rays, lab reports, and EHR text. Use when prototyping, fine-tuning, deploying, or validating MedGemma-based health AI under clinical data and safety controls.

--> <!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE --> --- name: 'test-time-knowledge-clinical-llm' description: 'Guide clinical LLM workflows that acquire and inject relevant medical knowledge at inference time to support safer decision-making without full fine-tuning.' measurable_outcome: 'Execute skill workflow successfully with valid output within 15 minutes.' allowed-tools: - read_file - run_shell_command - web_fetch ---

--> <!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE --> --- name: 'psychiatry-knowledge-fused-reasoning' description: 'Psychiatry CDS workflow for knowledge-fused augmented reasoning across diagnosis, risk assessment, medication reasoning, and care planning.' measurable_outcome: 'Execute skill workflow successfully with valid output within 15 minutes.' allowed-tools: - read_file - run_shell_command - web_fetch ---

Operate MedSAM2 for promptable segmentation of 3D medical images and medical videos, including CT lesion propagation, MRI volumes, RECIST-guided prompts, efficient CPU-oriented variants, training, and 3D Slicer integration. Use when generating or validating volumetric masks from sparse prompts or propagating masks through image slices or video frames.

--> --- name: memory-management description: Two-tier memory system that makes Claude a true workplace collaborator. Decodes shorthand, acronyms, nicknames, and internal language so Claude understands requests like a colleague would. CLAUDE.md for working memory, memory/ directory for the full knowledge base. keywords: - memory - context - decoding - glossary - collaboration measurable_outcome: Correctly identifies and decodes 90% of user-specific shorthands and acronyms in conversation. allo...

STATA code patterns for empirical accounting and finance research

Foundational Go patterns covering types, interfaces, goroutines, channels, and common idioms. Use when writing Go code, understanding Go's concurrency model, or needing guidance on which specialized Go skill to use. This is the entry point for Go development.

Elixir-specific library development patterns covering Hex package creation, OTP application design, Mix configuration, ExDoc documentation, typespecs, Dialyzer integration, and publishing best practices. Use when creating Elixir libraries, designing public APIs with OTP principles, managing dependencies, or publishing to Hex.pm. Extends meta-library-dev with Elixir ecosystem tooling.

Foundational Elixir patterns covering modules, pattern matching, processes, OTP behaviors (GenServer, Supervisor), Phoenix framework basics, and functional programming idioms. Use when writing Elixir code, building concurrent systems, working with Phoenix, or needing guidance on Elixir development patterns. This is the entry point for Elixir development.