last30days

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

• Third-party content exposure detected (high risk: 1.00). The skill's required workflow (SKILL.md Step 1–4) explicitly performs web_search/web_fetch, reddit_search (or site:reddit.com web searches), and bird (X/Twitter) searches to ingest Reddit posts, tweets, and web articles and then synthesizes quotes and action prompts—meaning untrusted, user-generated third-party content is read and used to drive decisions and outputs.