Claude Skills by anubhavg-icpl

Audit a content deployment's provenance chain across watermarking and C2PA metadata. Use when you need help with provenance audit.

Check a training dataset against California AB 2013 and EU TDM opt-out obligations. Use when you need help with provenance check.

Audit a function-calling integration against one provider for what breaks when ported to the other two. Use when you need help with provider portability audit.

Expert in the Publication design system - Print-inspired visual language for books, magazines, and reports with editorial grids and expressive typography. Use when building UI components, applying design tokens, or implementing visual styles for publication.

Build type-safe, validated LLM agents with Pydantic AI's RunContext, structured outputs, and dependency injection. Use when building AI applications with pydantic ai.

Pythonic code with modern type hints, dataclasses, async patterns, packaging, and testing. Use when you need help with python best practices.

Production-ready Python coding standards enforcing PEP 8, type hints, modern patterns, and maintainability. Use when enforcing python coding conventions and style rules.

python-expert. Use when writing, reviewing, or refactoring python code.

Production-ready Python project structure architect - validates and scaffolds enterprise-grade Python packages with src layout and modern tooling. Use when scaffolding, structuring, or architecting python projects.

Choose QA architecture, retrieval strategy, and evaluation plan. Use when you need help with qa architect.

qa-automation-expert. Use when writing, running, or improving tests with qa automation.

qa-specialist. Use when writing, running, or improving tests with qa.

Deep expertise in Qdrant — payload filtering, scalar/binary/PQ quantization, multi-vector, dense+sparse hybrid, and distributed mode. Use when implementing vector search, embeddings storage, or similarity queries with qdrant.

4-bit quantized LoRA fine-tuning with NF4, double quantization, and paged optimizers. Use when fine-tuning, training, or adapting language models with qlora techniques.

Pick between GGUF K/IQ quants, AWQ, GPTQ, bitsandbytes NF4, EXL2, MLX 4-bit, NVFP4 — decision matrix by hardware and serving stack. Use when creating, converting, or publishing model files with quantization format.

Pick a 2026 quantization format given hardware, engine, workload, and quality tolerance, and produce a calibration + validation plan. Use when you need help with quantization picker.

Expert in quantum computing fundamentals, Qiskit, and quantum algorithms. Use when you need help with quantum computing.

Query expansion, decomposition, multi-query retrieval, step-back prompting. Use when building or optimizing retrieval-augmented generation pipelines with query rewriting.

queue-stack-bloom. Use when you need help with queue stack bloom.

Configure a Qwen2.5-VL or Qwen3-VL deployment — resolution bounds, dynamic-FPS policy, window-attention flag, and JSON agent output mode — for a target video or image task. Use when you need help with qwen vl pipeline designer.

Expert in Qwik resumability, lazy execution, server functions, and Qwik City routing. Use when building web applications with qwik.

Expert R development for statistical computing, data science, and visualization. Use when writing, reviewing, or refactoring r code.

Expert in RabbitMQ message broker for reliable messaging. Use when configuring, deploying, or managing rabbitmq infrastructure.

Advise on SGLang adoption and prompt-ordering discipline for prefix-heavy workloads that want RadixAttention's cache reuse. Use when you need help with radix scheduler advisor.

RAGAS metrics, needle-in-haystack, RULER, golden sets, faithfulness and context precision/recall. Use when building or optimizing retrieval-augmented generation pipelines with rag eval.

Expert in Retrieval-Augmented Generation systems and knowledge bases. Use when you need deep expertise in rag.

RAGAS metrics for RAG and agent evaluation — faithfulness, relevancy, context precision/recall. Use when evaluating, monitoring, or observing LLM performance with ragas.

Expert in Railway services, environments, Postgres, and PR previews. Use when deploying to or building on railway edge/serverless platform.

Raj from The Big Bang Theory - astrophysicist who used to be unable to talk to women, hopeless romantic, and lover of all things fancy. Use when you need help with rajesh koothrappali.

RAPTOR — recursive abstractive summarization tree for long-document RAG. Use when building or optimizing retrieval-augmented generation pipelines with raptor.

Expert in Ray remote, actors, Ray Data, Ray Serve, and Ray Tune. Use when working with ray for data processing, streaming, or analytics.

Expert in the Raycast design system - Productivity launcher. Sleek dark chrome, vibrant gradient accents. Use when building UI components, applying design tokens, or implementing visual styles for raycast.

Design a relation extraction pipeline with provenance and canonicalization. Use when you need help with re designer.

Production-ready React coding standards enforcing component patterns, hooks, state management, and accessibility. Use when enforcing react coding conventions and style rules.

react-native-developer. Use when developing mobile applications with react native.

React 19 patterns including Server Components, Actions, Suspense, hooks, and component composition. Use when you need help with react patterns.

Expert in React Router 7 framework mode (formerly Remix) — loaders, actions, file routes. Use when building web applications with react router.

Production-ready React project structure validation and scaffolding with Vite, featuring modern patterns, TypeScript, and comprehensive tooling. Use when scaffolding, structuring, or architecting react vite projects.

Pick transport, VAD, streaming STT, LLM, streaming TTS, and orchestration for a target end-to-end latency. Use when you need help with realtime voice pipeline.

Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation

Recommend a red-team tool stack and configuration for a given deployment. Use when you need help with red team stack.

redis-expert. Use when designing, querying, or optimizing redis databases.

Redis patterns including caching strategies, pub/sub, streams for event processing, Lua scripts, and data structures. Use when you need help with redis patterns.

Expert in Redis caching, data structures, pub/sub messaging, persistence, clustering, performance tuning, and production best practices. Use when designing, querying, or optimizing redis databases.

Deep expertise in Redis Stack / RediSearch — vector index types (FLAT, HNSW, SVS-VAMANA), KNN + range queries, hybrid filter syntax. Use when implementing vector search, embeddings storage, or similarity queries with redis vector.

Jailbreak suites, garak, PyRIT, Promptfoo red-team — adversarial testing for LLM apps. Use when evaluating, monitoring, or observing LLM performance with redteam llm.

Expert in Redux state management for React applications with modern patterns, best practices, and performance optimization. Use when building applications with the redux framework.

refactoring-master. Use when you need help with refactoring master.

references. Use when you need help with references.

Configure an evaluator-optimizer (Self-Refine / CRITIC) loop given task, verifier availability, and iteration budget. Use when you need help with refine loop.