adsense-site-auditor

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface.
  • Ingestion points: The skill gathers evidence by crawling user-supplied external URLs, fetching robots.txt and sitemaps, and inspecting local repository contents (defined in SKILL.md under 'Gather evidence').
  • Boundary markers: Absent. The instructions do not define specific delimiters or include warnings to ignore instructions that may be embedded within the crawled external content.
  • Capability inventory: The agent utilizes its web-browsing/crawling tools and file-reading capabilities to ingest content from the target site or codebase.
  • Sanitization: Absent. Untrusted content from target websites is processed directly by the model to evaluate against the audit checklist.
  • [EXTERNAL_DOWNLOADS]: References official Google AdSense documentation.
  • The skill instructs the agent to use official Google Help and Policy domains (e.g., support.google.com) as the source of truth for its audit rules, which is a standard and safe practice for this domain-specific task.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 02:08 AM
Security Audit — agent-trust-hub — adsense-site-auditor