| A test plan exists, but no results |
At the final, results are expected |
| "Met" without evidence |
Counted as unmet |
| Guide and code version differ |
The assessed product becomes ambiguous |
| Empty remaining-risk section |
Incomplete analysis |
| Midterm feedback was ignored |
The finding–action loop is part of the process |
| Real secret / personal data in the repo |
Values must be synthetic |