code-review

To audit an existing codebase for security issues and architecture violations, provide a project path to trigger a parallel multi-agent review that surfaces findings across backend and Flutter code. This generates a validated findings report for intake processing.

Yes, you can review a specific feature path instead of auditing the whole codebase by providing a scoped path. The multi-agent discovery workflow adjusts to the defined scope, coordinating specialized agents to surface relevant quality problems and security findings.

Multi-agent code review works for finding quality problems by spawning sw-architect, security-engineer, and backend/flutter-engineer agents in parallel. This orchestration validates discovered findings against defined constraints and outputs a structured report for intake.

The best way to generate a findings report before deciding what to fix is to run a discovery workflow that orchestrates parallel reviews. It identifies security, architecture, and quality issues, feeding the structured output directly into intake for further processing.

Yes, multi-agent code review works with both backend and Flutter code by coordinating backend and flutter-engineer agents during the parallel discovery process. This ensures comprehensive coverage of architecture violations and security issues across your specific tech stack.