What problem does it solve?
Orchestrate end-to-end ingestion of clinical sources by coordinating discovery, download, normalization, topic inference, classification, and annotation to produce ready-to-use artifacts for downstream computable rules.
Core Features & Use Cases
- Deterministic, CLI-driven workflow for source intake: downloads, normalization to Markdown with frontmatter, topic initialization, and artifact generation.
- Idempotent operations across plan/implement/verify modes, with clear handling for open, authenticated, and manual sources.
- Production-ready pipeline for HI evidence synthesis, enabling downstream extraction, formalization, and concept vocabulary generation.
- On-demand resources management and strict CLI-only I/O to maintain reproducibility and traceability.
- Example use: ingest a set of open sources to produce sources/normalized files and a topics/<topic>/process/concepts.yaml for L2/L3 processing.
Quick Start
Plan and run an end-to-end ingest for a topic using rh-skills ingest plan <topic> followed by rh-skills ingest implement <topic>.