Tp4
High
- Category
- MCP Tool Poisoning
- Confidence
- 92% confidence
- Finding
- This finding indicates the skill advertises configuration and repair behavior while the observable content appears to focus on verification/reporting and credential/config inspection instead of performing the promised operations. Such divergence is dangerous because users may authorize the skill under false assumptions, and agents may access local credential sources or produce incomplete changes while claiming support for broader cloud administration tasks.
