Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 90% confidence
- Finding
- The skill declares sensitive operational requirements such as Python execution and a persistent token.json file, but does not declare explicit permissions for file access, file writes, or network use. That under-specification is dangerous because the skill handles OAuth credentials and mailbox/calendar data, so users and policy systems may not realize it can store tokens locally and communicate with Google APIs.
