← All commands
Development · /optimize
Optimize
Find measurable bottlenecks and improve performance without sacrificing correctness.
When to use it
Use this when performance or resource use has been measured and you need to reduce a bottleneck safely.
Useful inputs
- Benchmark or profiler output
- Workload and data size
- Correctness and memory limits
Instruction
Identify likely bottlenecks and how to measure them. Explain complexity, memory, and correctness effects.
Example input
Measurement: Listing 10,000 records takes 2.4 s. The query loads a category per row. Assess indexing and N+1 behavior.
Example result
Priority hypothesis: N+1 query behavior. Measure: Compare query counts in SQL logs. Candidate change: Load categories in one query. Acceptance: Same results and fewer queries; benchmark before and after with identical data.
Common pitfall
Do not choose code that merely looks faster; measure it and check memory cost and correctness.