What problem does it solve?
This Skill helps you reduce latency, CPU time, and allocations in Go services by mapping identified bottlenecks to the right optimization patterns after you’ve profiled.
Core Features & Use Cases
- Profile-first workflow: guides you to confirm whether the bottleneck is on-CPU vs off-CPU (e.g., pprof vs fgprof) before changing code.
- Optimization decision tree: uses benchmark-and-diagnose methodology to choose the correct technique for allocations, CPU hotspots, GC pressure, I/O waits, caching, and hot-path improvements.
- Review and remediation: supports architecture-level scanning and focused hot-path review to suggest improvements and targeted benchmarks that quantify gains.
- Common performance anti-patterns: covers frequent pitfalls such as logging in hot loops, reflection overhead, interface boxing, backing-array retention, and cache-locality mistakes.
- Use cases: performance code review, designing an optimization plan after pprof/benchmarks reveal the issue, and selecting benchmark methodology to validate improvements without regressions.
Quick Start
Ask an AI assistant to: profile my Go service first (pprof/fgprof as appropriate), then choose the best single optimization pattern for the detected bottleneck, implement one change, and verify the improvement with benchstat.
