AgentDB — semantic vector database search

★ 8.1 · vibe-coding

agentdb-vector-search is a Claude Code skill that implements semantic vector search using AgentDB's high-performance vector database with HNSW indexing and sub-millisecond response times under 100 µs. It supports document storage with embeddings, hybrid search combining vector similarity and metadata filtering, batch insert operations, and binary or scalar quantization for 4–32× memory reduction. The database is initialized via npx agentdb@latest CLI commands, and Claude Code integration is available through an MCP server exposing agentdb_query, agentdb_store, and agentdb_stats tools. According to the source benchmarks, queries run 150× to 12,500× faster than traditional solutions at scale. Ideal for teams building RAG pipelines, semantic search engines, or intelligent knowledge bases on Node.js 18+ with OpenAI embeddings or custom embedding models.