# Benchmark Suite

Benchmark Suite runs throughput and latency benchmarks, measures resource consumption, and detects performance regressions in agent clusters.

- subagent · `type=agent` · Claude Code, Codex, Cursor · ★ 850 repo stars
- Author: ruvnet
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/ruvnet/agentic-flow/tree/HEAD/.claude/agents/optimization/benchmark-suite.md
- Updated: 2026-10-26
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/agents/benchmark-suite

## What it does

The subagent handles structured benchmarking across latency, throughput, resource usage, and scalability. In distributed setups, it evaluates agent coordination, network topology, fault tolerance, and load balancing between nodes.

Runs can execute sequentially or concurrently. A full test run applies 30 seconds of warmup and 10 seconds of cooldown, cycling through iterations before reporting metrics. Regression analysis combines CUSUM change-point statistical detection with anomaly scoring. Baseline comparisons require prior historical runs to measure divergence.

## Install

### Claude Code (native platform)

**Manually: put the files in place**
- File: `.claude/agents/benchmark-suite.md` — For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

### Codex

**Manually: put the files in place**
- File: `.codex/agents/benchmark-suite.toml` — Codex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/

### Cursor

**Manually: put the files in place**
- File: `.cursor/agents/benchmark-suite.md` — Cursor also reads .claude/agents/

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/ruvnet/agentic-flow/tree/HEAD/.claude/agents/optimization/benchmark-suite.md

## When to use it
- Measure latency and processing throughput under sequential or parallel loads
- Assess fault tolerance, node coordination, and topology changes in distributed swarms
- Catch performance regressions using CUSUM change-point statistical algorithms
- Flag anomalous metric shifts against historical baselines via anomaly detection models

## Example prompts
- Execute parallel latency and throughput benchmark suites with a 30-second warmup period
- Test cluster fault tolerance and agent coordination under high concurrent load
- Compare current benchmark outputs against historical baselines using CUSUM regression checks

## Similar
- [benchmark-honestly](https://affpapa.org/en/nevibecode/skills/benchmark-honestly.md) — skill · ★ 541 repo stars
- [golang-benchmark](https://affpapa.org/en/nevibecode/skills/golang-benchmark.md) — skill · 41800 installs
- [performance-benchmark-specialist](https://affpapa.org/en/nevibecode/skills/performance-benchmark-specialist.md) — skill · 457 installs
- [benchmark](https://affpapa.org/en/nevibecode/skills/garrytan-benchmark.md) — skill · 409 installs
- [performance-engineer](https://affpapa.org/en/nevibecode/agents/voltagent-performance-engineer.md) — subagent · ★ 6505 repo stars
- [perf](https://affpapa.org/en/nevibecode/skills/perf.md) — skill · 899 installs

From the same set: [ruvnet/agentic-flow](https://affpapa.org/en/nevibecode/by/ruvnet/agentic-flow.md) — 76 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
