This skill helps Claude build ETL pipelines with CocoIndex, a real-time data transformation framework designed for AI workloads. It covers the full workflow: writing flows that pull from sources like S3 or local files, applying transformations like chunking and embedding, then pushing to vector databases or knowledge graphs. The incremental processing model means it only touches changed data, which matters when you're syncing large document collections. Use this when you need to set up embedding pipelines, extract structured data with LLMs, or build search indexes that stay in sync with your sources. The skill walks through dependency setup, environment config with API keys, and the specific coding patterns CocoIndex uses.
npx skills add https://github.com/davila7/claude-code-templates --skill cocoindex