staff-code-review

You can run a parallel code review by coordinating three subagents that examine a pull request simultaneously, each surfacing prioritized observations, questions, and ideas. The triad review aggregates results into a unified report for large Python projects.

A three-lens PR review deploys three senior reviewer subagents in parallel, each applying a different perspective to explore the codebase, issue tracker, and tests. They return prioritized lists of observations, questions, and ideas consolidated by severity.

Yes, this parallel AI reviewer is designed for mid- to large-scale Python projects with complex diffs and multiple stakeholders. It scales to handle large PRs and cross-team alignment by running context-aware exploration of codebase, issues, tests, and docs.

The best way to get prioritized observations is initiating a three-lens PR review on your current branch. Three subagents run in parallel, enforce a prioritized output format with severities, and consolidate observations, questions, and ideas into a unified report.

The parallel review process performs context-aware exploration beyond just the diff. It examines the codebase, checks the issue tracker, and reviews tests and documentation to surface comprehensive observations, questions, and ideas.

You should avoid using the three-lens approach for small, trivial pull requests where the overhead of running three parallel subagents outweighs the need for in-depth quality assessment. It is built for complex diffs and evolving codebases requiring high-quality reviews.