Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 86% confidence
- Finding
- The skill instructs the agent to execute a shell script (`scripts/telegram-auto-topic.sh`) and declares required binaries, but it does not declare explicit permissions governing shell/code execution. That creates a real least-privilege gap: a skill that can invoke shell commands against message-derived inputs should be explicitly permission-scoped so execution is visible and constrained.
