continuous-learning

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

• Third-party content exposure detected (high risk: 0.90). Yes—SKILL.md's "Step 2: Research Best Practices" explicitly directs the agent to "search the web", "review top results and incorporate relevant information", and add "References" with source URLs, which requires fetching and interpreting open/public third‑party content that can change the extracted skill's actions.