stress-test

stress-test is a skill for Claude Code from sgl-project/rbg. It costs 39 tokens per session (4,315 once invoked), scanned A, original, Apache-2.0.

A guide for measuring how an RBG controller performs under workload in a Kubernetes cluster. Kubernetes is a system for running containers, and kwok simulates workload pods without needing real computing resources.

In plain words
What is it for?
It helps run create, update, and delete tests, profile the controller with pprof, collect logs, and recommend resource, request-rate, and burst settings through manual or auto-tuned tests.
Why use it?
Realistic load tests can be expensive or difficult when many actual machines or GPU workloads are required.

Skill for Claude Code

Written for Claude Code: installed under .claude/. Also seen: names the AskUserQuestion tool.

Not installable on its own: it runs a file from its repository that does not travel with it. Clone the repository, or install whatever ships that file. The line is FAKE_NODE_COUNT=5 bash test/stress/scripts/setup-kwok.sh.

Install

Getting it into your agent

There is no command for this one: it runs only inside a plugin, and the catalogue could not identify which plugin ships it. The source is linked below.

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for stress-test

README.md
[![agentmods](https://agentmods.dev/badge/skills/sgl-project/rbg/stress-test.svg)](https://agentmods.dev/skills/sgl-project/rbg/stress-test)
Your own site
<a href="https://agentmods.dev/skills/sgl-project/rbg/stress-test"><img src="https://agentmods.dev/badge/skills/sgl-project/rbg/stress-test.svg" alt="Measured on agentmods" height="20"></a>
Per session 39 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 4,315 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00039 $0.04315
Opus 5 $0.00019 $0.02158
Sonnet 5 $0.00008 $0.00863
Haiku 4.5 $0.00004 $0.00432

Measured 6d ago against content hash 1620cdfb4406, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

stress-test scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- `helm`, `kubectl`, `go`, `curl` available in PATH
.claude/skills/stress-test/SKILL.md · 426 lines

How it starts

The opening of the file, as written. The whole thing — 426 lines — stays where its author put it; the contents beside it link to each section on GitHub.

RBG Controller Stress Test

This skill orchestrates a full stress test of the RBG controller. It supports two modes:

  • Manual mode: User defines all parameters, skill runs tests and outputs analysis report
  • Auto-tuning mode: User defines the workload scenario, skill runs multiple rounds and recommends optimal controller resource/QPS/burst configuration

Architecture

The stress test uses in-cluster kwok to simulate workload pods:

  • kwok controller + stage-fast deployed via Helm in the existing cluster
  • Fake nodes created with type=kwok label and kwok.x-k8s.io/node taint
  • RBG controller runs as a real Pod (via Helm) on real nodes with configurable resources
  • Stress test workload pods schedule to fake nodes and are simulated as Ready by kwok
  • This allows testing controller performance without needing real GPU/compute resources

Prerequisites

  • A Kubernetes cluster accessible via current kubeconfig
  • helm, kubectl, go, curl available in PATH
  • RBG CRDs and controller already deployed (or will be deployed by this skill)

Workflow

Phase 1: Ask User for Scenario

Use AskUserQuestion to gather parameters:

Question 1 - Test Mode:

  • Manual mode (user sets all params, get report)
  • Auto-tuning mode (auto-iterate to find optimal config)

Question 2 - Controller Resources (affects scheduling and memory):

  • 4c/8g (small)
  • 8c/16g (medium, recommended)
  • 16c/32g (large)
  • 32c/64g (extra-large)

Question 3 - Controller Parameters:

  • max-concurrent-reconciles: 5, 10, 20, 50
  • kube-api-qps: 20 (default), 50, 100, 200
  • kube-api-burst: 30 (default), 100, 200, 400

Question 4 - RBG Workload Template:

  • Roles per RBG: 2, 3, 5
  • LeaderWorkerPattern roles: 0, 1, 2
  • LWS size (pods per instance): 2, 4, 8
  • Total RBG instances: 10, 50, 100, 500, 1000

Question 5 - QPS for Operations:

  • Create QPS: 1, 5, 10, 20
  • Update QPS: 1, 5, 10, 20
  • Delete QPS: 1, 5, 10, 20

Question 6 - In-place Update:

  • Yes (update image to trigger InPlaceIfPossible)
  • No (update annotations to trigger RollingUpdate with RecreatePod)

Read the full file on GitHub · 426 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 6d ago First seen · 426 lines · 39 tokens per session scan A 1620cdfb4406

Subscribe to this mod's changes

stress-test is a skill published in the GitHub repository sgl-project/rbg (290 stars, last pushed yesterday), licensed Apache-2.0. It adds 39 tokens to every session and 4,315 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories