# Grounded RAG Ingestion

> - Chunk a document into provenance-carrying JSONL ready for embedding via the Nutrient Data Extraction API (/extraction/parse, spatial output). Every chunk carries element type, page index, bounding box, confidence, and reading order, traceable to a specific page region. Use for grounded or auditable RAG, cited retrieval, confidence-aware ingestion, or any pipeline that must prove which document region a retrieved answer came from. Triggers include grounded RAG, auditable RAG, RAG with provenance, chunk with bounding box, cite page, provenance-carrying chunks, confidence-aware chunking, or embedding pipeline. Not for cheap Markdown — use document-extraction-api with --output-format markdown for that. Not for known target fields — if the user wants specific fields ("invoice number and total", "map to my schema", "return these fields cited"), use document-extraction-api's extract (one cited call), not whole-document chunking.

## Facts
- Page: https://tashan.sh/capability/skill-pspdfkit-labs-grounded-rag-ingestion
- tashan id: skill:PSPDFKit-labs/grounded-rag-ingestion
- Source: https://github.com/PSPDFKit-labs/nutrient-skills
- Type: skill
- Category: search
- tashan score: not scored (catalogued only — too little public evidence)
- Adoption: 9.0
- Upkeep: 99.0
- Freshness: 97.0
- Evidence coverage: 84% of the inputs this score can use
- Health: active
- Instruction depth: not yet graded
- Official: no

## Install

```sh
cp -r grounded-rag-ingestion ~/.claude/skills/
```

## Security audit
Not scanned. We audit npm-published capabilities; this one has no npm package we can resolve, or has not reached the queue. This is not a clean bill of health.

---
Measured 2026-09-12 by tashan (https://tashan.sh) from public evidence. Scorer s5.