compare-flights
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill functions by ingesting and processing data from various external APIs and web search results, which introduces a vulnerability surface for indirect prompt injection.
- Ingestion points: Aggregates data from Duffel, Ignav, Seats.aero, and web search results (via Tavily or Exa) for platforms like Skiplagged and Kiwi.com.
- Boundary markers: The instructions do not specify any delimiters, boundary markers, or "ignore embedded instructions" warnings for the external data being integrated into the agent's context.
- Capability inventory: The skill orchestrates multiple tool calls and skill executions, including
duffel,ignav,seats-aero,chase-travel,amex-travel, andsouthwest(which may involve browser automation via Docker). - Sanitization: No explicit sanitization, validation, or filtering of the retrieved external data is described before the information is processed or displayed.
- [COMMAND_EXECUTION]: The skill directs the agent to execute a variety of tools and browser-automated tasks (specifically for portal logins and Southwest searches) to gather pricing and availability data.
Audit Metadata