Choose the right embedding model and chunking strategy for your vector search setup. Covers Voyage AI models (Anthropic's recommendation for Claude apps), OpenAI's text-embedding-3 series with Matryoshka dimension reduction, and local options like BGE and E5. Includes domain-specific models for code, finance, and legal documents. Templates handle token-based chunking, sentence boundaries, and semantic sectioning by headers. The model comparison table saves you from benchmarking common options yourself. Most useful when you're past the "just use whatever" phase and need to optimize retrieval quality or costs.
npx skills add https://github.com/wshobson/agents --skill embedding-strategies