marketing-agentic-search-optimizer
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill involves auditing external web pages and user journeys, which serves as an ingestion point for untrusted data.
- Ingestion points: Auditing of web pages, forms, and interactive elements across external sites during the readiness audit process.
- Boundary markers: Not present. The instructions do not specify the use of delimiters or clear separation between site content and agent instructions.
- Capability inventory: The agent generates reports and audits; the provided implementation templates demonstrate the use of standard web APIs like
fetchfor relative path requests. - Sanitization: Not present. No explicit validation or filtering of content from audited pages is described.
- [DATA_EXFILTRATION]: The skill includes a JavaScript template for dynamic action registration that uses
fetchto send data to a relative path (/api/bookings). This is standard behavior for web forms and does not involve sending data to unauthorized external domains.
Audit Metadata