doanchienthangdev/omgkit - 第 2 页

ai-agents

Building AI agents - tool use, planning strategies (ReAct, Plan-and-Execute), memory systems, agent evaluation. Use when building autonomous AI systems, tool-augmented apps, or multi-step workflows.

原文语言：英语

ai-architecture

AI application architecture - gateway, orchestration, model routing, observability, deployment patterns. Use when designing AI systems, scaling applications, or building production infrastructure.

原文语言：英语

ai-system-evaluation

End-to-end AI system evaluation - model selection, benchmarks, cost/latency analysis, build vs buy decisions. Use when selecting models, designing eval pipelines, or making architecture decisions.

原文语言：英语

dataset-engineering

Building and processing datasets - data quality, curation, deduplication, synthesis, annotation, formatting. Use when creating training data, improving data quality, or generating synthetic data.

原文语言：英语

evaluation-methodology

Methods for evaluating AI model outputs - exact match, semantic similarity, LLM-as-judge, comparative evaluation, ELO ranking. Use when measuring AI quality, building eval pipelines, or comparing models.

原文语言：英语

finetuning

Finetuning Foundation Models - when to finetune, LoRA, QLoRA, PEFT techniques, memory optimization, model merging. Use when adapting models to specific domains, reducing costs, or improving performance.

原文语言：英语

foundation-models

Understanding Foundation Models - architecture, sampling parameters, structured outputs, post-training. Use when configuring LLM generation, selecting models, or understanding model behavior.

原文语言：英语

guardrails-safety

Protecting AI applications - input/output guards, toxicity detection, PII protection, injection defense, constitutional AI. Use when securing AI systems, preventing misuse, or ensuring compliance.

原文语言：英语

inference-optimization

Optimizing AI inference - quantization, speculative decoding, KV cache, batching, caching strategies. Use when reducing latency, lowering costs, or scaling AI serving.

原文语言：英语

prompt-engineering

Designing effective prompts - system/user prompts, few-shot learning, chain-of-thought, defensive prompting, injection defense. Use when crafting prompts, improving outputs, or securing AI applications.

原文语言：英语

rag-systems

Retrieval-Augmented Generation - chunking strategies, embedding, vector search, hybrid retrieval, reranking, query transformation. Use when building RAG pipelines, knowledge bases, or context-augmented applications.

原文语言：英语

ai-engineering

Building production AI applications with Foundation Models. Covers prompt engineering, RAG, agents, finetuning, evaluation, and deployment. Use when working with LLMs, building AI features, or architecting AI systems.

原文语言：英语

user-feedback

Collecting and using user feedback - explicit/implicit signals, feedback analysis, improvement loops, A/B testing. Use when improving AI systems, understanding user satisfaction, or iterating on quality.

原文语言：英语

managing-databases

AI agent performs database administration tasks including backup/restore, monitoring, replication, security hardening, and maintenance operations. Use when managing production databases, troubleshooting performance, or implementing high availability.

原文语言：英语

managing-database-migrations

AI agent implements safe database migrations with zero-downtime strategies, rollback plans, and expand-contract patterns. Use when creating migrations, deploying schema changes, or implementing zero-downtime database updates.

原文语言：英语

designing-database-schemas

AI agent designs production-grade database schemas with proper normalization, indexing strategies, and data modeling patterns. Use when creating new databases, designing tables, modeling relationships, or reviewing schema architecture.

原文语言：英语

building-with-supabase

AI agent builds full-stack applications with Supabase PostgreSQL, authentication, Row Level Security, Edge Functions, and real-time subscriptions. Use when building apps with Supabase, implementing RLS policies, or setting up Supabase Auth.

原文语言：英语

developing-with-mongodb

The agent implements MongoDB NoSQL database solutions with document modeling, aggregation pipelines, and Mongoose ODM. Use when building document-based applications, designing schemas, writing aggregations, or implementing NoSQL patterns.

原文语言：英语

developing-with-prisma

The agent implements Prisma ORM for type-safe database access with schema design, migrations, and queries. Use when building database layers, designing relational schemas, implementing type-safe queries, or managing database migrations.

原文语言：英语

developing-with-redis

The agent implements Redis caching, data structures, and real-time messaging patterns. Use when implementing caching layers, session storage, rate limiting, pub/sub messaging, or distributed data structures.

原文语言：英语

deploying-to-aws

The agent implements AWS cloud solutions with Lambda, S3, DynamoDB, ECS, and CDK infrastructure as code. Use when building serverless functions, deploying containers, managing cloud storage, or defining infrastructure as code.

原文语言：英语

automating-with-github-actions

The agent implements GitHub Actions CI/CD workflows with builds, tests, and deployments. Use when setting up continuous integration, automating deployments, creating reusable actions, or implementing security scanning.

原文语言：英语

orchestrating-with-kubernetes

The agent implements Kubernetes container orchestration with deployments, services, Helm charts, and production patterns. Use when deploying containerized applications, configuring autoscaling, managing secrets, or setting up ingress routing.

原文语言：英语

profiling-performance

Performs browser performance profiling with Lighthouse, Core Web Vitals, and DevTools analysis. Use when auditing page performance, optimizing Core Web Vitals, analyzing bundle sizes, or implementing performance budgets.

原文语言：英语

building-django-apps

Builds enterprise Django applications with DRF, ORM optimization, async views, and Celery tasks. Use when creating Python web apps, REST APIs, or full-stack Django projects.

原文语言：英语

building-express-apis

Builds production Express.js APIs with TypeScript, middleware patterns, authentication, and error handling. Use when creating Node.js backends, REST APIs, or Express applications.

原文语言：英语

building-fastapi-apis

Builds high-performance FastAPI applications with async/await, Pydantic v2, dependency injection, and SQLAlchemy. Use when creating Python REST APIs, async backends, or microservices.

原文语言：英语

building-laravel-apis

Builds enterprise Laravel applications with Eloquent, API Resources, Sanctum auth, and queue processing. Use when creating PHP backends, REST APIs, or full-stack Laravel applications.

原文语言：英语

building-nestjs-apis

Builds enterprise NestJS applications with TypeScript, dependency injection, TypeORM, and microservices patterns. Use when creating scalable Node.js backends, REST/GraphQL APIs, or microservices.

原文语言：英语

building-react-apps

Builds production React applications with hooks, TypeScript, state management, and performance optimization. Use when creating React SPAs, component systems, or interactive UIs.

原文语言：英语

building-vue-apps

Builds production Vue 3 applications with Composition API, TypeScript, Pinia, and Vue Router. Use when creating Vue.js SPAs, component libraries, or reactive web interfaces.

原文语言：英语

implementing-accessibility

Claude implements WCAG 2.1 AA compliant web interfaces with proper ARIA, keyboard navigation, and screen reader support. Use when building accessible components or auditing existing UIs.

原文语言：英语

designing-frontend-patterns

Claude designs scalable React component architectures using compound components, custom hooks, and state machines. Use when building reusable UI systems or complex component APIs.

原文语言：英语

building-responsive-layouts

Claude builds responsive web layouts with mobile-first CSS, fluid typography, and container queries. Use when creating adaptive UIs that work across all device sizes.

原文语言：英语

building-with-shadcn

Claude builds accessible React UIs using shadcn/ui components with Radix primitives and React Hook Form integration. Use when creating forms, dialogs, or composable UI systems.

原文语言：英语

styling-with-tailwind

Claude styles UIs with Tailwind CSS utility classes, responsive design, and dark mode support. Use when creating component styles, theming systems, or responsive layouts.

原文语言：英语

building-3d-graphics

Claude builds immersive 3D web experiences with Three.js and React Three Fiber. Use when creating WebGL scenes, 3D animations, shaders, or physics simulations.

原文语言：英语

writing-javascript

Writes modern JavaScript with ES2024+ features, async patterns, and functional programming. Use when building Node.js applications, implementing async workflows, or writing clean maintainable code.

原文语言：英语

brainstorming-ideas

AI agent generates diverse solutions through structured divergent thinking and systematic exploration frameworks. Use when exploring options, solving problems creatively, or generating alternatives.

原文语言：英语

implementing-defense-in-depth

AI agent designs layered security architecture with multiple independent protective barriers ensuring no single point of failure. Use when building security systems, reviewing architecture, or hardening applications.

原文语言：英语