rollinggo-hotel-booking

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (low risk: 0.10). The required workflow queries hotel descriptions and details via the rgh CLI tool interacting with the RollingGo hotel booking service, which retrieves third-party hotel information that can contain outsider-authored text and descriptions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs and executes downloads of executable binaries and npm packages from external GitHub repositories and registries during setup, presenting weak/unknown/suspicious operational dependencies.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 0.70). The skill implements a hotel booking assistant via CLI (rgh) with direct financial operations including price confirmation, order placement ("book"), and cancellations/refunds. Step 0 mentions checking status or logging in via rgh login, but there is no documented service authentication or validation enforced independently by the service/execution layer for booking or payment routes—only agent-level workflows and CLI parameters. Therefore, it is a direct financial operation without established qualifying independent access control.