Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 90% confidence
- Finding
- The skill instructs the agent to read multiple reference files and run a local script, which implies file_read and file_write/code-adjacent capabilities, but no permissions are declared. This creates a trust and policy gap: the runtime may allow filesystem access that reviewers and users cannot see, increasing the risk of unauthorized local file access or modification if the skill is reused in a broader environment.
