Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 95% confidence
- Finding
- The skill declares capabilities indicating access to environment variables and outbound network use, but no explicit permissions model is declared. That creates a real security gap because the skill can read sensitive configuration such as WOLFRAM_APP_ID and transmit user-provided data to an external API without clear user-visible authorization boundaries. In a networked third-party API skill, this is more dangerous than a purely local utility because queries may contain sensitive data and the missing permission declaration reduces transparency and control.
