Automated review is putting a change through an AI reviewer before a person sees it. A second agent, or a tool built for the job, reads the diff and comments: possible bugs, missed edge cases, style drift, security smells. It is the machine version of a code review, run as a first pass.
What it is good at
Unlike an automated check, which only knows pass or fail, an automated review can reason about the change and explain itself. It is tireless and consistent. It never skims because it is Friday afternoon, and it catches a lot of the obvious stuff: an unhandled null, a leaked secret, a function that quietly grew to two hundred lines. Clearing that noise before a human looks means the human spends their attention on what actually needs judgment.
It is best understood as a relative of self-critique: a fresh model attacking work it did not write, with no attachment to the approach.
- Run it on every change as a cheap first filter.
- Expect false positives. It flags things that are fine, so weigh its comments.
- It has no real stake in the outcome and no memory of your intent.
Related terms
Automated check
An automated check is a machine-verifiable gate that agent output has to pass, like tests, a type check, a linter, or a build. It either passes or fails with no judgment, which makes it the backbone of trusting agent code, especially when you are running unattended.
Read definition →Human review
Human review is a person actually reading what an agent produced, understanding it, and taking responsibility for shipping it. It is the final quality gate that tests and automated review can support but never replace.
Read definition →Self-critique
Self-critique is asking a model to attack its own output, or having a fresh agent do it, to catch bugs and bad assumptions before you ship. It is a direct counter to a model's tendency to agree with whatever it just produced.
Read definition →