Tp4
- Category
- MCP Tool Poisoning
- Confidence
- 96% confidence
- Finding
The documented purpose says the skill transforms prior work into reusable process assets, but the detected behavior indicates local artifact validation, integrity/path checks, and secret-pattern scanning that are not disclosed. This mismatch is dangerous because users and policy systems may authorize the skill for one low-risk purpose while it performs broader inspection of local data, increasing the chance of unintended data exposure or overreach.
- Content
