puppeteer

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

• Third-party content exposure detected (high risk: 0.85). This skill’s runtime workflow uses Puppeteer to navigate to user-supplied URLs and then ingests rendered page text/DOM via page.evaluate() / page.content() (and may also ingest JSON from page.on('response')), which is outsider-authored content from arbitrary third-party web pages/APIs.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (medium risk: 0.60). The prompt explicitly instructs disabling Chrome sandboxing (e.g., --no-sandbox, --disable-setuid-sandbox) and recommends bypass/stealth techniques, which instructs the agent to bypass security mechanisms and thus increases risk to the host.