multi-review

Parallel code reviews across multiple domains are orchestrated by dispatching specialist sub-agents simultaneously for a given diff. This process applies to non-trivial diffs touching security, cloud-functions, frontend, and edge-cases, producing a final synthesized report.

The best way to get a unified code review report is through multi-agent orchestration that deduplicates findings from parallel specialist reviewers. It generates a greptile-style final structured verdict with guidance to drive safe merges for production-bound changes.

Yes, you can use multi-review for production-bound diffs that require multi-domain scrutiny. It coordinates niche specialist agents to perform parallel reviews and provides a final structured verdict to ensure safe merges for non-trivial changes.

To start an orchestrated code review, invoke /multi-review on a non-trivial diff. This initiates parallel specialist reviews across domains and returns a single synthesized report with deduplicated findings and risk-based verdict formatting.

Yes, parallel code review deduplicates findings from different specialist agents. The orchestration process consolidates overlapping issues identified by domain specialists into a single, unified greptile-style final report.

You should use orchestrated parallel code reviews when dealing with non-trivial diffs that touch multiple domains like security, cloud-functions, frontend, and edge-cases. Standard reviews lack the parallel specialist execution and synthesized risk-based verdict needed for complex changes.