Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 94% confidence
- Finding
- The skill instructs the agent to run bundled Python code, execute shell commands, and perform file reads and writes against an arbitrary repository, yet it declares no permissions. This creates a transparency and policy-enforcement gap: an operator or host system may treat the skill as lower-risk than it really is, while the actual workflow can modify files in place and access potentially sensitive repository contents.
