Missing User Warnings
Medium
- Confidence
- 95% confidence
- Finding
- The skill explicitly documents a paid execution path (`/v1/run`) that spends Orthogonal credits, but it does not provide a clear user-facing warning that invoking the tool causes billable actions. In an agent setting, this can lead to unintended charges if the model autonomously follows the documented workflow and triggers paid API calls without informed user consent.
