design-benchmark

design-benchmark is a command for Claude Code from Owl-Listener/ai-design-skills. It costs 12 tokens per session (491 once invoked), scanned A, original, MIT.

A plan for a benchmark suite: a fixed collection of representative tests used to track an AI product over time.

In plain words
What is it for?
Use it to select common cases, edge cases, and known failure cases; define quality metrics and baselines; and set monitoring thresholds.
Why use it?
It makes changes such as model or prompt updates easier to compare and helps reveal regressions, meaning new problems caused by a change.

Command for Claude Code

Written for Claude Code: argument-hint in frontmatter.

Part of the evaluation plugin — 7 skills, 3 commands shipped together

Good fit Use it to select common cases, edge cases, and known failure cases; define quality metrics and baselines; and set monitoring thresholds.

Compare 6 commands from other repositories ↓
Install with agentmods
npx agentmods add commands/owl-listener/ai-design-skills/design-benchmark
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo
git clone --depth 1 https://github.com/Owl-Listener/ai-design-skills

Made for: Claude Code.

Or install evaluation, the plugin that ships this one along with the rest of its 7 skills, 3 commands.

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 design-benchmark

README.md
[![agentmods](https://agentmods.dev/badge/commands/owl-listener/ai-design-skills/design-benchmark/github.svg)](https://agentmods.dev/commands/owl-listener/ai-design-skills/design-benchmark)
Your own site
<a href="https://agentmods.dev/commands/owl-listener/ai-design-skills/design-benchmark"><img src="https://agentmods.dev/badge/commands/owl-listener/ai-design-skills/design-benchmark/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for design-benchmark

Your own site · 80×15
<a href="https://agentmods.dev/commands/owl-listener/ai-design-skills/design-benchmark"><img src="https://agentmods.dev/badge/commands/owl-listener/ai-design-skills/design-benchmark.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 12 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 491 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 $0.00012 $0.00491
Opus 5 $0.00006 $0.00246
Sonnet 5 $0.00002 $0.00098
Haiku 4.5 $0.00001 $0.00049

Measured 13d ago against content hash b4ccad99cdef, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

design-benchmark scanned grade A with 0 findings 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 13d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

claude-plugin/evaluation/commands/design-benchmark.md · 51 lines

How it starts

The opening of the file, as written. The whole thing — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are designing a benchmark suite for ongoing AI quality measurement. Use only skills from the evaluation plugin. Follow this process:

Step 1: Define Benchmark Goals

  • What does this benchmark need to measure?
  • What decisions will benchmark results inform? (model updates, prompt changes, feature launches)
  • What's the measurement cadence? (daily, weekly, per-release)

Step 2: Design the Golden Test Set

Using output-quality-rubrics and longitudinal-measurement:

  • Define input categories that cover key use cases
  • Create 20-50 representative test inputs spanning:
    • Common use cases (high frequency)
    • Edge cases (high risk)
    • Known failure modes (regression detection)
    • Diverse user profiles and contexts
  • For each input, define the expected quality standards

Step 3: Define Metrics

Using task-success-metrics, user-satisfaction-signals, and output-quality-rubrics:

  • Select metrics to track: quality scores, task success rates, failure rates, latency
  • Define how each metric is calculated
  • Set baselines from current performance
  • Define acceptable ranges and alert thresholds

Step 4: Design Failure Monitoring

Using failure-taxonomy:

  • Define which failure types to track
  • Set severity-specific thresholds (e.g., zero tolerance for critical safety failures)
  • Design failure trend tracking

Step 5: Design Comparative Testing

Using comparative-evaluation:

  • Define how new versions will be compared to the current baseline
  • Specify A/B test protocols for significant changes
  • Define statistical significance requirements

Step 6: Plan for Drift Detection

Using longitudinal-measurement:

  • Define what constitutes meaningful quality drift vs. normal variance
  • Design automated alerts for drift detection
  • Specify the investigation and response protocol when drift is detected

Output

Deliver a complete benchmark suite specification:

  1. Benchmark goals and decision framework
  2. Golden test set (20-50 inputs with expected standards)
  3. Metrics definitions with baselines and thresholds
  4. Failure monitoring specifications
  5. Comparative testing protocol
  6. Drift detection and response plan
  7. Benchmark reporting template
  8. Recommended automation approach

Read the full file on GitHub · 51 lines

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. 13d ago First seen · 51 lines · 12 tokens per session scan A b4ccad99cdef

Subscribe to this mod's changes

design-benchmark is a command published in the GitHub repository Owl-Listener/ai-design-skills (173 stars, last pushed 3mo ago), licensed MIT. It adds 12 tokens to every session and 491 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.