stealth-scraper

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes multiple examples and templates that embed proxy credentials verbatim into URLs, environment variables, and code (e.g., http://USER:PASS@host, HTTP_PROXY, requests proxies), which would require an agent to insert real usernames/passwords directly into generated commands or configs.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This skill explicitly provides step-by-step methods to evade anti-bot protections (residential proxy rotation, TLS/browser fingerprint spoofing, webdriver/property tampering, human-behavior simulation, CAPTCHA solving and sticky sessions) and is clearly intended to facilitate large-scale, likely unauthorized web scraping and data collection — high abuse potential though it contains no obvious backdoor/RCE or direct credential-theft routines.

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

• Third-party content exposure detected (high risk: 1.00). Yes — SKILL.md explicitly instructs the agent to scrape arbitrary public websites (e.g., "Scrape any website" and code examples using the browser tool and curl_cffi like requests.get(url) and browser.evaluate to navigate/scroll/extract pages), so the agent will fetch and interpret untrusted third-party web content that could contain actionable instructions.