summarize

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

• Third-party content exposure detected (high risk: 0.95). The skill’s required workflow ingests outsider-authored free text by fetching and extracting content from user-supplied URLs (public web pages) or YouTube links at runtime, which is then summarized by the agent’s LLM context.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill fetches arbitrary external URLs at runtime (e.g., "https://example.com" and "https://youtu.be/dQw4w9WgXcQ") and injects the fetched page/transcript content into the model prompt for summarization, which can directly control the agent's instructions via remote content.