This study examines 1.02 million reviewed pull requests from 207 GitHub projects across three code-review eras: human-centric review, LLM-assisted review, and agentic review. It identifies Gradual AI Adoption, Rapid LLM Adoption, and Rapid AI Agent Adoption. Agent-involved patterns, especially reviews initiated by AI agents or involving multiple agents, are associated with faster review decisions under gradual adoption and rapid agent adoption. However, faster decisions do not correspond to better review quality. Once AI reviewers participate, human-AI interaction patterns become the strongest explanatory factor for review efficiency, while review activity and pull-request type remain important across eras.
No heat snapshots are available in the last 24 hours.