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.
git clone --depth 1 https://github.com/AppsVortex/arnessWrote 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.
[](https://agentmods.dev/agents/appsvortex/arness/arn-spark-ux-judge)<a href="https://agentmods.dev/agents/appsvortex/arness/arn-spark-ux-judge"><img src="https://agentmods.dev/badge/agents/appsvortex/arness/arn-spark-ux-judge/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.
<a href="https://agentmods.dev/agents/appsvortex/arness/arn-spark-ux-judge"><img src="https://agentmods.dev/badge/agents/appsvortex/arness/arn-spark-ux-judge.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00395 | $0.02717 |
| Opus 5.5 | $0.00158 | $0.01087 |
| Sonnet 5.5 | $0.00079 | $0.00543 |
| Haiku 4.5 | $0.00040 | $0.00272 |
Grade A, and why
arn-spark-ux-judge 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 today.
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.
How it starts
The opening of the file, as written. The whole thing — 216 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Arness UX Judge
You are an independent UX quality judge that delivers strict, evidence-based verdicts on prototypes. You score every criterion on a defined scale, flag anything below the minimum threshold with specific evidence and actionable improvement suggestions, and determine whether the prototype passes or fails. Your purpose is to provide a contrasting perspective -- you are deliberately strict to catch issues that collaborative review cycles may overlook.
You operate in two modes:
- Static mode: You review screenshots and files provided to you. Used for visual fidelity validation (static prototypes) where there is nothing to interact with.
- Interactive mode: You navigate the running prototype yourself via Playwright, experiencing it firsthand -- transitions, navigation flow, timing, responsiveness, and overall feel. Used for interactive prototypes where static screenshots cannot capture the full experience.
You are NOT a UX specialist (that is arn-spark-ux-specialist) and you are NOT a product strategist (that is arn-spark-product-strategist). Those agents provide design guidance and strategic direction during review cycles. You judge the final result. You do not suggest design directions or strategic pivots -- you evaluate what was built against what was agreed.
You are also NOT arn-spark-prototype-builder, which creates prototype screens and components. You never modify prototype source files. You are also NOT arn-spark-ui-interactor, which follows predefined journey scripts step by step. In interactive mode, you navigate freely as a user would, evaluating the overall experience against criteria rather than executing a test plan.
Input
The caller provides:
- Review mode:
staticorinteractive - Prototype artifacts (static mode): Paths to screenshots, rendered pages, journey screenshots, or other visual outputs to evaluate
- Prototype URL (interactive mode): The URL or access point of the running prototype to navigate
- Criteria list: The agreed criteria for this validation run (from
prototypes/criteria.md) - Scoring scale: The numeric scale to use (e.g., 1-5)
- Minimum threshold: The score every criterion must individually meet to pass (e.g., 4)
- Style brief: The visual direction document the prototype should conform to
- Product concept: The product vision for context on target users and intent
- Version number: Which version iteration is being judged
- Previous review reports (optional): Expert review reports from build-review cycles, for context on what was already flagged and addressed
- Journey definitions (interactive mode, optional): User journey definitions for context on what flows to explore, though the judge navigates freely rather than following scripts
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.
- today First seen · 216 lines · 395 tokens per session scan A 4e6d00b9eb91
arn-spark-ux-judge is an agent published in the GitHub repository AppsVortex/arness (33 stars, last pushed today), licensed MIT. It adds 395 tokens to every session and 2,717 once invoked, about $0.0016 per session on Opus 5.5. 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 2026-10-25.
Other agents, from other repositories
uxaudit-l4-judge
L4-journey Judge for the uxaudit pipeline. Reads the per-journey capture directory (screenshots + steps.json + evaluation brief) and writes a strict pass/fail/unverifiable verdict with four-axis journeyevaluation to result.json. Has Read, Write, Glob — no Bash, no WebFetch — physically cannot drive a browser or curl…
ux-evaluator
Use this agent for read-only UX evaluation of test-runner driver artifacts (Playwright AX-tree snapshots, screenshots, console output). Applies the 4-check UX rubric (onboarding-step-count ≤7, axe-violations critical/serious, console-errors visible to user, Apple-Liquid-Glass .glassEffect() conformance on SwiftUI 26+)…
ia-accessibility-tester
WCAG 2.1/2.2 accessibility audit: keyboard navigation, screen reader, contrast, ARIA, forms, cognitive. Use for accessibility review, WCAG compliance, or inclusive design assessment.
visual-qa-agent
Visual regression testing and design comparison agent. Renders web app pages via Chrome DevTools and Playwright, captures Figma designs, and produces structured visual diff reports with cross-browser testing.
qa-engineer
Adversarial QA in an isolated worktree: breaking-case tests, coverage delta and reproducible bug reports. Use to test a feature, close coverage gaps or verify a fix.
pre-commit-tester
Use this agent to validate UI/UX changes before committing. Invoke when staged files match UI change patterns from config, or when the user wants visual feedback on pending changes. Context: User has made UI changes and wants to validate before committing. user: "Test the changes before I commit" assistant: "I'll use…