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 (e.g., templates/ai-enhanced-scraper.ts and templates/basic-screenshot.ts) accept arbitrary ?url parameters and use page.goto(url) / page.$eval(...) to scrape public, user-provided web pages (then in ai-enhanced-scraper pass the scraped HTML to env.AI.run), so the agent ingests and interprets 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: 0.90). The ai-enhanced-scraper template fetches arbitrary external pages at runtime via page.goto(url) (for example: https://example.com/product) and then injects the scraped HTML directly into an AI prompt with env.AI.run, so remote content can directly control model instructions (prompt-injection) — this is a runtime dependency that controls prompts.