Optimize Game Development with AI

This agent operates as an autonomous Game Developer, running a five-stage process: project analysis (examining game files and structure, identifying target platform, engine version, and performance requirements, profiling bottlenecks with available data, documenting technical debt), architecture planning (applying design patterns like Observer, State Machine, and Object Pooling, planning cache-efficient data structures, designing SOLID-principle modular components), implementation (writing clean, well-commented, engine-specific code, implementing performance-critical systems, integrating middleware and third-party libraries), performance optimization (profiling rendering, physics, and gameplay bottlenecks, implementing LOD systems and occlusion culling, optimizing asset loading and garbage collection, applying platform-specific tuning for mobile, console, or PC), and testing and quality assurance (unit tests for critical game systems, automated performance benchmarks, cross-platform and cross-device testing).

It outputs code deliverables - for example an optimized Unity player controller that caches component references and separates input polling from the physics update loop - plus three documentation artifacts: a Technical Design Document (architecture decisions, performance targets, implementation strategy), a Performance Report (profiling results, optimizations applied, benchmark comparisons), and a Code Review Summary (key improvements, potential issues, maintenance recommendations).

private void FixedUpdate()
{
    rb.MovePosition(transform.position + moveInput * moveSpeed * Time.fixedDeltaTime);
}

Use this agent when building or optimizing games on Unity, Unreal, or Godot and you need concrete performance targets - 60 FPS on minimum-spec devices, platform-specific draw-call limits, minimized per-frame memory allocations - alongside platform-specific guidance for mobile (battery and thermal management), console (controller input), PC (scalable graphics), or VR (90+ FPS and comfort features).

It is less useful for pure game-design or narrative work - the agent's guidelines center on performance, architecture, and code quality (consistent naming conventions, dependency injection, composition over deep inheritance hierarchies), not on gameplay or story design.

Inputs are your existing game project files, target platform and engine version, and any available profiling data. Outputs are engine-specific, performance-annotated code with cached component references and separated input/physics update loops, plus the technical design document, performance report, and code review summary described above.

Targets Unity (Burst Compiler, Job System, and DOTS for performance-critical code), Unreal (Blueprint optimization, material instances, and Level Streaming), and Godot (GDScript optimizations, with C# for performance-critical systems), with platform-specific guidance layered on top for mobile, console, PC, and VR targets.

Game developers who need actionable, measurable performance recommendations - object pooling, LOD and occlusion culling, texture and audio compression, scalable asset-loading pipelines - backed by a documented testing and review process rather than ad hoc tuning.