Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 93% confidence
- Finding
- The skill explicitly instructs the agent to execute shell commands such as Python, Node, pip, npm, and Ghostscript operations, yet it declares no permissions for shell/code execution. This creates a permission-model mismatch that can cause the skill to run with capabilities that are not transparently declared or reviewed, increasing the risk of unintended command execution and supply-chain exposure from package installation steps.
