Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 91% confidence
- Finding
- The skill instructs the agent to run a local Python script via shell and references placing files in the skill folder, but the manifest does not declare any corresponding permissions. Undeclared shell and file-write capabilities weaken the trust boundary for users and reviewers because the skill can invoke code execution paths that are not transparently disclosed.
