Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 89% confidence
- Finding
- The skill instructs the agent to execute shell commands (`source`, `lasutil`, `jq`, file writes/uploads) but does not declare any permissions or capability boundaries. Undeclared shell access is dangerous because users and reviewers may not realize the skill can access local files, environment variables, and networked resources, increasing the chance of unintended command execution or data exposure.
