Select 20 papers from the papers table with the highest citations_count that currently have no outgoing kg_edges. For each paper: (1) extract 3–5 key empirical claims using LLM; (2) identify relevant entities (genes, pathways, diseases) in canonical_entities; (3) create kg_edges linking the paper's findings to these entities. Use edge_type values like 'supports', 'identifies', 'demonstrates'. This enriches the knowledge graph with structured evidence from high-impact literature.
Last Error
cli-get-next: phantom running task — linked run c152ed5f-489 already terminal (abandoned); requeued immediately