seo

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs the agent to fetch and analyze public webpages and Google SERPs (e.g., "How Claude does keyword research" in 3.9 and "Competitor & SERP Analysis" in 3.10 which say to "WebFetch the competitor's page" and "WebFetch Google SERP"), so the agent will read and act on untrusted third‑party content from arbitrary websites.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs the agent to use WebFetch at runtime to fetch external pages (e.g., Google SERP pages such as https://www.google.com and competitor URLs) and then extract and inject that fetched content into the model's analysis/context, which directly influences prompts and outputs.