메뉴
Route RAG chunking decisions across semantic, hierarchical, sliding-window, contextual-header, and framework-selection strategies.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use this skill when building, debugging, or improving Retrieval-Augmented Generation systems, including chunking, vector database selection, hybrid search, reranking, multimodal RAG, code documentation RAG, retrieval latency, and production RAG architecture.
Chunk nested documents into parent-child levels so retrieval can move from broad sections to fine-grained passages.
Use semantic boundaries and embedding similarity to chunk text for higher-relevance retrieval.
Use overlapping windows to preserve context across chunk boundaries while controlling retrieval size.
Reduce retrieval latency with caching, batching, and index-level optimization.
Route RAG performance work for latency, caching, indexing, filtering, batching, and query optimization.
| name | chunking |
| title | Chunking |
| description | Route RAG chunking decisions across semantic, hierarchical, sliding-window, contextual-header, and framework-selection strategies. |
| category | chunking |
| tags | ["chunking","routing","rag","preprocessing"] |
| allowed-tools | ["Read","Grep","Glob"] |
Use this parent skill when the main RAG problem is how to split source material into retrievable units. Route to the child skill that best matches the document shape, retrieval failure, and production constraints.
Poor chunking causes retrieval misses, fragmented answers, duplicated context, and weak citations. RAG systems need chunking strategies that preserve meaning while staying efficient for indexing and retrieval.
Use semantic chunking for prose, hierarchical chunking for structured documents, and code-specific guidance for APIs or source repositories.
Use sliding windows when answers cross boundaries, contextual headers when chunks lack source context, and semantic chunking when fixed windows split ideas.
Measure retrieval recall, answer groundedness, duplicate context rate, and citation usefulness before changing embeddings or vector databases.
