Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 92% confidence
- Finding
- The skill documents shell command capabilities and expected CLI execution (`etf-assistant ...`) but does not declare corresponding permissions. Undeclared execution capability is dangerous because it can bypass normal trust and review expectations, and if the implementation invokes shell commands with user-controlled inputs such as ETF codes or search strings, it could expand into command execution or data access risks.
