Qodo solutions engineer Cole Gannaway describes code reviewCode review examines proposed source changes to assess their correctness, clarity and suitability before or around integration. as a bottleneck when AI agents produce changes faster. Using an illustrative data-model change, he explains how a single line may affect clients and repositories beyond the file being edited.
He discusses a software graph and custom rules as ways to provide reviewers and coding agents with relevant repository contextContext engineering designs how relevant instructions, evidence, state and tool information are assembled for a model's task.. He also argues for independent reviewIndependent evaluation assesses a system through reviewers whose judgment is sufficiently separate from the organization developing or promoting it., potentially using different models or harnessesAn AI agent harness is the software framework that packages a model with tools, instructions, context management, execution controls, and user interaction., rather than having the implementing agent assess its own work alone.
During questions, Gannaway discusses connecting reviews to the intent of a task and using periodic checks before a pull request. He presents Qodo’s product as one implementation of these ideas rather than a guarantee that all errors will be caught.
Watch on YouTube
