Backend Developer • Go, Infrastructure, Caching
Your high partial success rate indicates deep engagement with complex infrastructure problems.
You're doing foundational infrastructure work—database migrations, caching layers, queue workers. These are inherently complex tasks.
When you provide file paths and narrow scope, you succeed. Specificity is your leverage point.
The AI often struggles with Go project structure discovery. Provide known paths upfront:
In cmd/queueworker/main.go and internal/cache/redis.go, I need to migrate Redis to Valkey.
Your exploration and multi-task prompts have higher trouble scores. Reframe broad requests:
When migrating, point to working implementations:
We have a working Valkey connection in internal/cache/valkey.go. Use that pattern.
You're tackling the hardest backend work—database migrations, caching systems, queue workers.
You rescued 4 sessions independently when the AI went off track. Strong engineering judgment.