web-scraper
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill implements a robust 'Output Sanitization Layer' specifically designed to detect and remove prompt injection vectors (such as hidden HTML comments and zero-width characters) from untrusted web content before it is processed by an LLM.
- [DATA_EXFILTRATION]: The skill includes a 'URL Validation Layer' that prevents Server-Side Request Forgery (SSRF) and data exfiltration by blocking private IP ranges and dangerous URI schemes like data: and javascript:.
- [REMOTE_CODE_EXECUTION]: To mitigate risks of arbitrary code execution, the skill mandates a 'Library-Only Mode' and explicitly disables crawl4ai hooks, which are noted as potential RCE vectors in specific CVE contexts.
- [COMMAND_EXECUTION]: The skill uses dynamic context injection (!cat) to load local protocol and configuration files. These operations are restricted to non-sensitive project paths and are used for environment initialization.
Audit Metadata