Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 92% confidence
- Finding
- The skill instructs the agent to read cloud credentials from environment variables and invoke a script that performs outbound network actions, but the skill metadata declares no permissions. This mismatch is a real security issue because it hides the skill's effective access to secrets and external services, reducing transparency and preventing proper policy review before a high-impact action like bulk calling is triggered.
