Sustainability RSS Fetch

Core Goal

• Ingest all RSS/Atom items into SQLite before topic filtering.
• Use doi as the primary key in entries.
• Keep RSS metadata isolated in its own DB file.
• After semantic screening, keep relevant rows and prune non-relevant rows to DOI-only.

Triggering Conditions

• Receive a request to import sustainability feeds and persist all fetched records first.
• Receive a request to do prompt-based topic screening after DB ingestion.
• Receive a request to convert irrelevant rows into lightweight DOI-only records.
• Need stable DOI-keyed storage for downstream API/fulltext/summarization.

Mandatory Workflow

1. Prepare runtime and RSS metadata DB path.