team-polish

Coordinate a cross-disciplinary polish sprint by orchestrating role-based subagents across performance, visual, audio, and QA phases. This workflow hardens a feature before release, ensuring cross-team alignment and quality without bottlenecks through a defined six-phase process.

The best way to run visual and audio polish in parallel is using a phased workflow that executes Phase 3 (Visual Polish) and Phase 4 (Audio Polish) simultaneously. This accelerates delivery while maintaining comprehensive quality across multiple subsystems.

Error recovery during a polish workflow surfaces blockers and offers retry, skip, or stop options. If a blocker cannot be resolved, the protocol generates a partial report, ensuring robust error handling across the cross-disciplinary sprint.

Roles in a comprehensive QA and polish workflow include performance-analyst, engine-programmer, technical-artist, sound-designer, tools-programmer, and qa-tester. These role-based subagents drive the phased orchestration to achieve a polished release.

Yes, you can use this workflow for performance tuning and technical art refinement. It is applicable to performance analysis, visual polish, audio refinement, and comprehensive QA, coordinating the relevant subagents across these subsystems.

If a blocker occurs during the final readiness assessment, the error recovery protocol surfaces the issue and offers retry, skip, or stop options. It can generate a partial report to document the current state of the polish sprint.