cloudflare-browser-rendering

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

• Third-party content exposure detected (high risk: 1.00). The skill's templates and documentation (e.g., SKILL.md Quick Start and templates/ai-enhanced-scraper.ts, templates/basic-screenshot.ts, templates/web-scraper-basic.ts) explicitly accept arbitrary ?url parameters, navigate to and scrape public websites' HTML, and even pass that untrusted page content to an AI extractor — meaning the agent ingests and acts on third-party user-generated web content that could contain indirect prompt-injection payloads.

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

• Potentially malicious external URL detected (high risk: 1.00). The templates (templates/ai-enhanced-scraper.ts) fetch an arbitrary user-supplied URL at runtime (e.g., https://example.com/product via ?url=...) with puppeteer, extract the page HTML, and inject that content directly into env.AI.run messages, so remote content fetched at runtime can directly control the model prompt.