I am not a game developer but a platform developer. The custom reference approach is spot on and that's a massive job.
I compiled 202 tips from 10 'important' CC users into a repo and plugin. You might want to ask CC to review it and see if any of it is worth implementing in your skills as well: https://github.com/4riel/cc-bible
At that speed you can treat it as disposable: fresh instance per test run, no shared state, no flaky tests from leftover S3 objects. that was never practical with LocalStack cold start
53% catch rate on bugs that already cleared human review, with the main bottleneck being false positives... Sounds about right for where LLMs are useful: they don't get tired on patch 47 of a series