RAG Orchestration Skills
Skills that orchestrate the full RAG pipeline: question → retrieval → BigQuery semantic layer → structured answer.
What this is
The semantic layer is a set of BigQuery tables enriched with LLM-generated categories, topics, embeddings, and sentiment scores. Skills in this repo embed a user's question, retrieve the most relevant context, and synthesize a grounded answer — no hallucination, sources cited.
Current skills:
- cx-rag-researcher — Customer Experience: SUMO/Kitsune forums, Zendesk tickets, Mozilla Knowledge Base
Install
Quick example
What are users saying about Firefox sync in Q1 2026?
Claude embeds the question, retrieves context from the CX semantic layer, and returns a grounded answer — no fabrication, sources cited.