icp-website-audit
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements a standard auditing workflow. It relies on user-provided URLs and writes its analysis to a structured local directory. No indicators of credential theft, data exfiltration, or malicious persistence were found.\n- [PROMPT_INJECTION]: The skill ingests untrusted content from external websites, which is a common surface for indirect prompt injection. 1. Ingestion points: Website crawling in Phases 2 and 3. 2. Boundary markers: Not explicitly defined. 3. Capability inventory: File system write access, web fetch, and web search. 4. Sanitization: Relies on standard agent tool boundaries.
Audit Metadata