hum-source-criticism

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

• Third-party content exposure detected (high risk: 1.00). The skill's SKILL.md and references (e.g., references/fact-check-tools.md and the LinkedIn example) explicitly instruct the agent to search, fetch, and evaluate public web pages and social-media posts (e.g., a LinkedIn post, techinsight-asia.com, Snopes/PubMed, reverse-image searches), meaning it will read untrusted third-party content that can influence its decisions and actions.