---
description: Extract entity-level claims from text for StemeDB ingestion
---

Extract entity-level claims from the provided text using the `extract-claims` skill.

**Usage:** `/extract-claims "Your text or paste content here"`

**What this does:**
1. Analyzes the text for all mentioned entities (explicit and implied)
2. Extracts a SEPARATE claim for EACH entity mentioned
3. Identifies implicit claims (category membership, relationships)
4. Assigns confidence scores based on language hedging
5. Returns structured JSON ready for StemeDB API

**Example:**
```
/extract-claims "Every mainstream database, from PostgreSQL to MongoDB to Neo4j, enforces single value per key."
```

This would extract 7 claims:
- 4 storage_model claims (one per database + category)
- 3 is_mainstream claims (implicit category membership)

**For CLI usage:**
```bash
# From text
npx tsx community/scripts/extract-claims.ts --text "Your text" --source-class Expert

# From file and submit
npx tsx community/scripts/extract-claims.ts --file article.txt --source-class Clinical --submit

# Dry run from stdin
cat paper.txt | npx tsx community/scripts/extract-claims.ts --stdin --dry-run
```

Load the `extract-claims` skill for full extraction methodology.