repurpose

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

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md explicitly accepts arbitrary web URLs and uses the WebFetch tool to "Fetch the content" (see "If URL provided: Fetch the content using WebFetch" and "URL Input" sections) and then embeds that fetched "ORIGINAL CONTENT" into the generate_content prompt, so untrusted third-party page content could supply instructions that materially influence generation.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly uses WebFetch at runtime to fetch arbitrary URLs (e.g., https://example.com/blog-post) and then injects the fetched content into the generate_content instructions, so remote content can directly control the agent's prompts.