Claude Skills by anubhavg-icpl

Efficient storage and retrieval of genomic variant data using TileDB. Scalable VCF/BCF ingestion, incremental sample addition, compressed storage, parallel queries, and export capabilities for population genomics. Use when you need help with tiledbvcf.

Guide for building Graph Neural Networks with PyTorch Geometric (PyG). Use this skill whenever the user asks about graph neural networks, GNNs, node classification, link prediction, graph classification, message passing networks, heterogeneous graphs, neighbor sampling, or any task involving torch_geometric / PyG. Also trigger when you see imports from torch_geometric, or the user mentions graph convolutions (GCN, GAT, GraphSAGE, GIN), graph data structures, or working with relational/network...

Generate concise (3-4 page), focused medical treatment plans in LaTeX/PDF format for all clinical specialties. Supports general medical treatment, rehabilitation therapy, mental health care, chronic disease management, perioperative care, and pain management. Includes SMART goal frameworks, evidence-based interventions with minimal text citations, regulatory compliance (HIPAA), and professional formatting. Prioritizes brevity and clinical actionability. Use when you need help with treatment p...

UMAP dimensionality reduction. Fast nonlinear manifold learning for 2D/3D visualization, clustering preprocessing (HDBSCAN), supervised/parametric UMAP, for high-dimensional data. Use when you need help with umap learn.

Access comprehensive LaTeX templates, formatting requirements, and submission guidelines for major scientific publication venues (Nature, Science, PLOS, IEEE, ACM), academic conferences (NeurIPS, ICML, CVPR, CHI), research posters, and grant proposals (NSF, NIH, DOE, DARPA). This skill should be used when preparing manuscripts for journal submission, conference papers, research posters, or grant proposals and need venue-specific formatting requirements and templates. Use when you need help wi...

Use this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like \\"the xlsx ...

Chunked N-D arrays for cloud storage. Compressed arrays, parallel I/O, S3/GCS integration, NumPy/Dask/Xarray compatible, for large-scale scientific computing pipelines. Use when you need help with zarr python.

RFC compliance validator for SCIM Provisioning Gateway implementations - SCIM schema/protocol, Bearer Token, request routing. Use when you need help with scim provisioning gateway validator.

Configure Stable Diffusion / Flux inference for a given prompt, style, and quality bar. Use when you need help with sd prompter.

Compose ControlNets, LoRAs, and IP-Adapters on top of an SD / Flux base for a given set of inputs. Use when you need help with sd toolkit composer.

SD3 / SD3.5 Large, MMDiT architecture, T5-XXL prompting, differences from SDXL. Use when working with multimodal AI (images, audio, video) using sd3.

Stable Diffusion XL - base + refiner, LoRA, IP-Adapter, samplers, schedulers. Use when working with multimodal AI (images, audio, video) using sdxl.

Pick a search strategy (ReAct, ToT, LATS, evolutionary) given task shape, token budget, and evaluator quality. Use when you need help with search policy.

Expert in secret management with HashiCorp Vault, AWS Secrets Manager, and secure secrets handling. Use when performing security analysis, auditing, or hardening with secret management.

Application security covering input validation, auth, headers, secrets management, and dependency auditing. Use when you need help with security hardening.

Use this skill when adding authentication, handling user input, working with secrets, creating API endpoints, or implementing payment/sensitive features. Provides comprehensive security checklist and patterns.

Expert in security testing methodologies and penetration testing. Use when writing, running, or improving tests with security testing.

Audit a proposed self-improvement or constitutional AI pipeline before it runs at scale. Use when you need help with self improvement auditor.

Self-RAG — model decides when to retrieve and critiques evidence with reflection tokens. Use when building or optimizing retrieval-augmented generation pipelines with self rag.

GPTCache, Helicone cache, LangChain semantic cache — embedding-based dedup for LLM apps. Use when evaluating, monitoring, or observing LLM performance with semantic cache.

Expert in the Sentry design system - Error monitoring. Dark dashboard, data-dense, pink-purple accent. Use when building UI components, applying design tokens, or implementing visual styles for sentry.

Choose encoder-decoder vs decoder-only for a new sequence-to-sequence task. Use when you need help with seq2seq picker.

Pick sequence architecture (RNN, transformer, SSM, hybrid) given length, throughput, and training budget. Use when you need help with sequence architecture picker.

Use after() for Non-Blocking Operations. Use when you need help with server after nonblocking.

Cross-Request LRU Caching. Use when you need help with server cache lru.

Per-Request Deduplication with React.cache(). Use when you need help with server cache react.

Parallel Data Fetching with Component Composition. Use when you need help with server parallel fetching.

Minimize Serialization at RSC Boundaries. Use when you need help with server serialization.

Expert in serverless architecture with AWS Lambda, Azure Functions, and Google Cloud Functions. Use when you need deep expertise in serverless.

Expert in AI engineering development environments, tooling, and workflows from the AI Engineering from Scratch curriculum. Use when you need help with setup and tooling.

Supervised fine-tuning fundamentals — chat templates, packing, completion-only loss, NEFTune. Use when fine-tuning, training, or adapting language models with sft techniques.

Serve LLMs with SGLang's RadixAttention, structured outputs (compressed FSM), tensor parallel, DP-attention, and PD disaggregation. Use when deploying, running, or configuring local LLM inference with sglang.

Expert in the Shadcn design system - Shadcn/ui-inspired design with minimal, clean components, monochrome palette, and utility-first patterns. Use when building UI components, applying design tokens, or implementing visual styles for shadcn.

Expert in shadcn/ui copy-paste pattern, registry system, CLI 3.0, and theming. Use when building web applications with shadcn.

Expert in shadcn/ui component library, Radix primitives, Tailwind CSS styling, and accessible React components. Use when you need help with shadcn ui.

Adult Sheldon Cooper from The Big Bang Theory - theoretical physicist, genius, and master of social awkwardness with strict routines and roommate agreements. Use when you need help with sheldon cooper.

Expert in the Shopify design system - E-commerce platform. Dark-first cinematic, neon green accent, ultra-light type. Use when building UI components, applying design tokens, or implementing visual styles for shopify.

Plan a sim-to-real transfer pipeline for a given robot + task, covering DR, SI, and safety. Use when you need help with sim2real planner.

Simon Willison — Datasette, LLM CLI, building in public, plugin architectures, daily practical AI. Use when you want code review, architecture advice, or opinions in the style of simon willison.

|. Use when you need help with simple deck.

Expert in the Simple design system - Straightforward, no-frills design with clean typography, neutral colors, and intuitive layouts that stay out of the way. Use when building UI components, applying design tokens, or implementing visual styles for simple.

Simple Preference Optimization — reference-free, length-normalized preference alignment. Use when fine-tuning, training, or adapting language models with simpo techniques.

Design a generative-agent simulation (Smallville-style) for a given scenario. Specifies memory schema, reflection cadence, plan horizon, spatial/social constraints, and evaluation metrics. Use when you need help with simulation designer.

Expert in the Skeumorphism design system - Real-world mimicry with textured surfaces, 3D effects, and familiar physical metaphors for intuitive digital interfaces. Use when building UI components, applying design tokens, or implementing visual styles for skeumorphism.

Pick the right 3DGS export format (.ply / .splat / glTF KHR_gaussian_splatting / USD) given the downstream viewer or engine. Use when you need help with skill 3dgs export router.

Build production-grade RAG with hybrid search, reranking, and evaluation. Use when you need help with skill advanced rag.

Given a dataset of ground-truth boxes, run k-means on (w, h) and return anchor sets per FPN level plus coverage statistics. Use when you need help with skill anchor designer.

Choose the right anomaly detection approach for your problem. Use when you need help with skill anomaly detector.

Build, debug, and reason about automatic differentiation systems. Use when you need help with skill autodiff.

Establish a strong classification baseline before reaching for complex models. Use when you need help with skill classification baseline.