Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 96% confidence
- Finding
- The skill instructs the agent to use shell, file reads/writes, environment-backed secrets, and network access, but it does not declare those permissions. That creates a trust and review gap: a host may expose more capability than users or policy expect, and the skill can send outbound notifications and write local config containing sensitive credentials.
