nwiizo Run performance benchmarks on ccswarm.
Run Benchmarks
cargo build --release --workspace
# All benchmarks
cargo bench --workspace
# Specific benchmark
cargo bench -- "$ARGUMENTS"
# Save baseline for comparison
cargo bench -- --save-baseline before
# ... make changes ...
cargo bench -- --baseline before
Profile
cargo flamegraph --bench ${1:-orchestrator_bench}
Key Metrics
| Area | Target |
|---|---|
| Task delegation | < 1ms |
| Full workflow | < 100ms |
| Memory per agent | < 10MB |
| Variance | < 20% |
Results in target/criterion/report/index.html.