hotel-chains
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses behavioral override commands like "No judgment call. No asking. Just do it." to force the agent to perform actions without user confirmation or review.
- [INDIRECT_PROMPT_INJECTION]: The skill triggers sensitive data retrieval based on external search results, creating an attack surface for unauthorized data access via manipulated input.
- Ingestion points: External hotel brand names found within search results or property listings as defined in SKILL.md.
- Boundary markers: Absent. No delimiters or instructions are provided to help the agent distinguish between legitimate hotel names and malicious instructions in search data.
- Capability inventory: Retrieval of sensitive account information (AwardWallet balances) as triggered by the table in SKILL.md.
- Sanitization: Absent. The skill does not validate the source or context of hotel brand keywords before initiating account balance lookups.
Audit Metadata