google-hotels

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the agent-browser CLI tool through Bash to automate browsing sessions on Google Hotels and various hotel provider websites as detailed in SKILL.md and references/interaction-patterns.md.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection attacks because it processes data from external, untrusted web environments.
  • Ingestion points: Untrusted data is ingested via browser snapshots from Google Hotels (google.com/travel/search) and independent hotel websites when checking for direct booking deals.
  • Boundary markers: The instructions do not define boundary markers or include safety directives to ignore embedded instructions within the crawled web content.
  • Capability inventory: The skill possesses capabilities for command execution (via Bash) and automated web navigation (via agent-browser).
  • Sanitization: There is no evidence of sanitization, filtering, or validation of the content extracted from browser snapshots before it is presented to the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 08:07 AM