microsite
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and fetches font assets from well-known Google hosting services (fonts.googleapis.com and fonts.gstatic.com).
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from multiple external enrichment and search tools to dynamically populate landing page templates, creating a surface for potential prompt injection via external content.
- Ingestion points: External data enters the agent context through tools such as
enrich_company,enrich_person, andsearch_knowledge_baseas defined inreferences/octave-tool-reference.md. - Boundary markers: The skill utilizes structured markdown and HTML templates (
references/microsite-outline-template.md,references/html-architecture.md) to define output segments. - Capability inventory: The skill is capable of generating full HTML files containing CSS and inlined JavaScript, and it can perform network operations via enrichment tools.
- Sanitization: The
references/task-method.mdfile mandates that the agent inspect all final claims, quotes, and proof points, and verify social metadata and URL destinations after deployment.
Audit Metadata