Claude Skills by aicodedecode

RabbitMQ guidance — exchanges, queues, routing, acknowledgments, dead-lettering, clustering, and operations.

Radio astronomy fundamentals — interferometry concepts, spectral lines (HI, CO), calibration, and imaging basics.

Engineer retrieval-augmented generation systems — chunking, embeddings, hybrid retrieval, reranking, and grounded generation with citations. Use when building or fixing RAG pipelines.

RAG guidance — chunking, retrieval tuning, hybrid search, reranking, citations, evaluation, and production RAG.

Ruby on Rails guidance — MVC conventions, ActiveRecord mastery, Hotwire, background jobs, and production ops.

Master Raycast on macOS: commands, extensions, AI, clipboard history, window management, and custom scripts. Use when turning Raycast into the central command layer of a Mac workflow.

Design role-based access control — role engineering, permission modeling, hierarchies, and governance.

Idiomatic React: component design, hooks discipline, state placement, rendering performance, and modern patterns. Use when writing or reviewing React code, or choosing React architecture.

Build performant React forms with React Hook Form: validation, controlled inputs, dynamic fields, and error UX. Use when forms need speed and clean validation.

Professional React Native: cross-platform architecture, navigation, native modules, and performance. Use when writing, reviewing, or structuring React Native apps.

Build agents that interleave reasoning and acting — the ReAct loop of thought, action, and observation.

Write professional React: hooks discipline, composition, performance, error boundaries, and modern patterns. Use for any React development beyond basics.

Manage server state with TanStack Query: queries, mutations, caching, invalidation, optimistic updates, and pagination. Use when fetching/mutating remote data in React.

Master Readwise: highlight capture, spaced-repetition review, and exporting knowledge into your notes. Use when you read a lot digitally and want highlights to become lasting knowledge.

Writing effective rebuttals to peer review — point-by-point responses, tone, new analyses, and knowing when to concede.

Manage React state with Recoil: atoms, selectors, async selectors, and atom families/effects. Use when maintaining Recoil codebases or needing its data-flow graph.

Plan and execute authorized red-team operations that emulate real adversaries to test detection and response, with strict scoping and safety controls.

Red-team LLM systems methodically — scoping, adversarial test design, vulnerability classification, and remediation tracking. Use when you need to find weaknesses before attackers do. Defensive security practice only.

Redis guidance — data structures, caching patterns, pub/sub, streams, transactions, and production operations.

Redis data structures, caching strategies, and production operations — use when adding caching, queues, or rate limiting with Redis.

Manage complex state with Redux Toolkit: slices, thunks, RTK Query, selectors, and DevTools. Use for large apps needing strict state conventions.

Systematic refactoring: safe transformations, strangler patterns, characterization tests, and large-scale change management. Use when improving code structure without changing behavior.

Write, test, and debug regular expressions with pattern recipes, engine quirks, and performance safety.

Threat-model regulatory and compliance risk — mapping legal obligations to security controls and residual exposure.

Understand and apply reinforcement learning — MDPs, value and policy methods, exploration, reward design, and RLHF basics. Use when decisions are sequential and feedback comes as rewards.

Monitoring release health — crash-free rates, adoption, and catching bad deploys early — platform-agnostic patterns.

Run reliable releases: versioning, branching, checklists, staged rollouts, and rollback plans. Use when cutting releases or designing a release process.

Evaluating measurement quality — test-retest, inter-rater agreement, and construct validity evidence.

Build a reliable reminder setup with Muse: capture-everything habit, due dates vs ticklers, recurring reminders, follow-up checks, and escalation when things slip.

Idiomatic Remix: loaders/actions, progressive enhancement, nested routing, and web-standard patterns. Use when writing, reviewing, or structuring Remix applications.

Satellite and airborne remote sensing — multispectral indices, SAR basics, classification, and change detection.

Thrive in remote roles with async communication habits, home-office setup, and visibility strategies.

Assessing wind, solar, and hydro resources — measurement, modeling, uncertainty, and bankable energy-yield estimates.

Ship ML models as APIs with Replicate — run open models, deploy custom models, and pay per use.

Rerank retrieval candidates with cross-encoder models — two-stage pipelines, model selection, latency trade-offs, and evaluation. Use when first-stage retrieval is broad but imprecise.

Do applied AI research engineering — reproduce papers, run ablations, manage experiments, and turn prototypes into reliable systems. Use when bridging the gap between a research idea and working code.

Work with academic papers — find them, read efficiently, extract claims and methods, compare across papers, and track what matters. Use when doing literature-driven research with AI assistance.

Write academic papers with AI assistance — outlining, drafting sections, tightening arguments, managing citations, and responding to reviewers. Use when turning research into a publishable manuscript.

REST API design guidance — resource modeling, versioning, pagination, auth, error formats, and documentation.

Resting-state fMRI analysis — denoising, seed-based and ICA networks, and reliability-aware interpretation.

Turning a rejected grant into a funded one — triaging reviews, the introduction to resubmission, and strategic revision.

Extracting structured data from resumes — fields, skills, and normalization — use when building hiring pipelines.

Draft, tailor, and polish resumes for specific roles using achievement-based bullets and recruiter conventions.

Market for retail — in-store experience, promotions, merchandising, omnichannel, and foot traffic growth.

Reverse-engineer binaries and protocols for defensive purposes — vulnerability research, malware understanding, and interoperability.

Manage online reviews — generation strategies, response playbooks, reputation monitoring, and turning reviews into growth.

Reviewing scientific code — correctness, reproducibility, numerical soundness, and constructive feedback.

Build the full reinforcement-learning-from-human-feedback stack — preference data, reward model, and policy optimization.

Bundle libraries and apps with Rollup: plugins, tree-shaking, multi-format output, code splitting, and watch mode. Use when building distributable JS libraries.

Eliminate rote work: identify repetitive tasks, measure their cost, and automate or remove them systematically. Use when drowning in repetitive busywork or building a culture of leverage.