screenshot-gemini

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 open an arbitrary initial URL and navigate within that page (using a provided page-name and description) to capture content, which requires fetching and interpreting public/untrusted web pages at runtime and therefore could allow indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime command ("bunx @willbooster/agent-skills@latest screenshot ...") fetches and executes the npm package @willbooster/agent-skills@latest at runtime, so external code from that package (npm:@willbooster/agent-skills@latest) directly executes and is a required dependency.