Разработка LLM-приложений и RAG-систем

★ 6.8 · general

ai-engineer is a Claude Code skill that builds production-ready LLM applications, advanced RAG systems, and intelligent agents — from simple chatbots to complex multi-agent workflows. It covers the full stack: semantic chunking, embedding model selection, hybrid vector search with reranking across Pinecone, Weaviate, Chroma, and Qdrant, plus agentic loops with tool use and multi-model routing between fast and high-quality models. Enterprise concerns are built in — token cost tracking, latency monitoring, caching, A/B testing for prompt versions, prompt injection guards, and PII filtering. The skill is designed to be used proactively whenever an LLM feature, AI-powered application, or retrieval-augmented generation pipeline is involved. Ideal for engineers shipping customer support bots, knowledge-base assistants, or production AI services.