Agent Performance Benchmarker

This is a benchmarking framework for distributed consensus protocols that measures throughput, latency, and resource usage across different algorithms like Byzantine, Raft, and Gossip. It runs configurable test scenarios with adaptive load generation, collects time-series metrics, and generates comparative analysis between protocols. You'd reach for this when you need hard numbers to choose between consensus mechanisms or tune parameters for production workloads. The implementation handles ramp-up testing and includes percentile latency tracking, which matters when you're trying to avoid tail latency surprises. It's fairly heavyweight, so this is more for serious distributed systems work than quick proofs of concept.

Install to Claude Code

npx -y skills add ruvnet/ruflo --skill agent-performance-benchmarker --agent claude-code