longbridge-graham-screener

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

• Third-party content exposure detected (high risk: 1.00). The skill's workflow explicitly falls back to WebSearch (SKILL.md "WebSearch fallback" table and references/criteria.md dynamic-adjustment rules) to fetch public publisher URLs (PMI, sector outlook, insider transactions) whose findings are read and used to adjust scores and trigger the value‑trap override, and every WebSearch hit is required in the Data Source Appendix (references/output.md) — meaning untrusted third‑party content is ingested and can materially change decisions.