domestic-trip-planning
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external travel and booking websites via
web_searchandbrowser_navigate. This creates a surface for indirect prompt injection where instructions hidden in flight or hotel descriptions could influence agent behavior. - Ingestion points: Data is ingested via
web_searchandbrowser_navigateas described inSKILL.mdandreferences/web-extract-fallback-strategies.md. - Boundary markers: Explicit boundary markers for external data are not mentioned in the instructions.
- Capability inventory: The skill uses web search and browser navigation tools to retrieve information.
- Sanitization: No specific sanitization or filtering logic is defined for the ingested web content.
- [DYNAMIC_EXECUTION]: The skill includes JavaScript snippets in
references/ihg-hotel-scraping.mdfor DOM traversal to extract hotel pricing from IHG.com and interaction patterns inreferences/live-flight-pricing.md. These are static templates for data extraction within the browser environment and do not involve dynamic code generation from untrusted sources.
Audit Metadata