Glossary

Last updated 2 October 2026.

What do these code review terms mean?

Short, plain definitions of the terms used across the MergeSafe docs and in its reviews, each with a link to the page that goes deeper.

• AI code review

  An AI reads a code change and comments on bugs, security problems and risks before it merges. What it does well and where people still decide.

• Pull request review

  The check a proposed change gets before it merges: reading the diff, commenting on lines, and approving or asking for changes.

• Merge readiness

  How safe a pull request is to merge, as one number. MergeSafe scores it 1 to 5 from the severity of its verified findings.

• P0 and P1 severity

  The two highest severities a finding can have: P0 critical, such as a security hole or data loss, and P1 high, such as a feature that does not work.

• False positive

  A finding that reports a problem which is not real. Why they cost trust, how they are measured, and what happens when you dispute one.

• SAST vs AI code review

  Static analysis matches code against known patterns; AI review reads a change for what it means, logic bugs included. What each catches and why to use both.

• Review recall and precision

  Recall is how many real bugs a review finds; precision is how many of its findings are real. How to measure both, with MergeSafe's own numbers.