Raises test coverage by identifying uncovered code regions, ranking them by risk, and generating targeted tests that hit them — prioritizing branches and conditions over raw line count. Use when coverage is below target, when untested code is blocking a release, or when deciding which tests to write next.