ai-visibility-audit

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to install additional agent capabilities from the vendor's repository using 'npx skills add unifapi-agent/agents'. This is consistent with the skill author's infrastructure.
  • [PROMPT_INJECTION]: The workflow involves an indirect prompt injection surface where the agent uses the 'browser/markdown' tool to read and analyze content from external, third-party URLs.
  • Ingestion points: External URLs cited in AI answers are fetched and rendered as Markdown via 'browser/markdown' in SKILL.md.
  • Boundary markers: The instructions do not define clear delimiters or warnings to ignore embedded instructions within the fetched external content.
  • Capability inventory: The skill uses tools for SEO/GEO data retrieval and browser-based content reading (unifapi, browser/markdown).
  • Sanitization: There is no explicit requirement for the agent to sanitize or filter the content retrieved from external sources before processing it for the diagnostic pillars.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 10:28 AM
Security Audit — agent-trust-hub — ai-visibility-audit