Version 2.2.4

Benchmark workspace refresh

• Reworked the active-run dashboard with clearer run identity, status, progress, live throughput, latency, data-rate cards, output controls, and a terminal-style live log.
• Refined benchmark history and result presentation with denser summaries, clearer metric cards, and improved chart axes, smoothing, fills, markers, and stacked-series rendering.

Per-run benchmark outputs

• Benchmark CSV, JSON, and log artifacts now write to a folder named after the run ID so repeated runs no longer overwrite earlier results.
• Applied the per-run output layout consistently across the Python, Go, Rust, Java, Android, and mock benchmark implementations.