find-signal-leads

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The runtime path is scripts/match_entities.py reading outsider-authored free text from candidates.jsonl (a user/provider-supplied JSONL dataset via an external file/queue), then applying name-matching over the ingested {"name": ...} values without filtering for a specific pre-selected item.