melech-scout

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow performs multiple web searches across lanes (e.g., Canon/Ecosystem/Verdicts) and then ingests the resulting live pages/text as evidence into candidate rows, which can include outsider-authored free text (like HN/Reddit or arbitrary web content) posted by external users.