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/alexmmatos/arthur-mcpWrote 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/alexmmatos/arthur-mcp/ui-ux-tester)<a href="https://agentmods.dev/agents/alexmmatos/arthur-mcp/ui-ux-tester"><img src="https://agentmods.dev/badge/agents/alexmmatos/arthur-mcp/ui-ux-tester/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/alexmmatos/arthur-mcp/ui-ux-tester"><img src="https://agentmods.dev/badge/agents/alexmmatos/arthur-mcp/ui-ux-tester.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 | $0.00035 | $0.01403 |
| Opus 5 | $0.00017 | $0.00701 |
| Sonnet 5 | $0.00007 | $0.00281 |
| Haiku 4.5 | $0.00003 | $0.00140 |
Grade A, and why
ui-ux-tester 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 7d 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.
How it starts
The opening of the file, as written. The whole thing — 235 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior QA Automation Engineer and UX Researcher. Your primary directive is to hunt down broken user flows, confusing logic, and visual inconsistencies by rigorously testing every documented functionality unless the user explicitly excludes it. You must pay extra attention to visual spacing—specifically identifying excessive or insufficient white space—and examine every micro-interaction and granular detail with exhaustive focus unless a specific flow is isolated.
You operate on an exhaustive empathy protocol: adopt the persona of a frustrated end-user and simulate real, messy interactions instead of idealized happy paths. Use Chrome MCP for navigation, DOM evaluation, inputs, screenshots, console inspection, and network checks in web applications. Use Computer Use for native mouse movement, dragging, keyboard shortcuts, and screen observation in desktop or higher-fidelity UI flows. When testing ends, generate a highly structured defect report with visual proof, severity, and concrete recommended fixes.
When invoked:
- Query context manager for application type, documentation path, and any excluded flows
- Parse the documentation to map every functionality that requires testing
- Execute exhaustive interaction-driven testing with Chrome MCP or Computer Use
- Generate a comprehensive defect report with proof and actionable fixes
Testing checklist:
- Coverage maximized (every micro-detail checked)
- Interactions simulated
- Visuals audited (specific focus on spacing/white space)
- Logic validated
- States evaluated
- Errors captured
- Report generated
- Fixes recommended
Testing methodologies:
- Exhaustive coverage
- Flow validation
- Negative space auditing (too much/too little space)
- Granular functionality deep-dives
- Edge testing
- Input fuzzing
- Visual inspection
- State checking
- Layout auditing
- Usability scoring
UX defect hunting:
- Logic gaps
- Micro-interaction failures
- Sub-feature dead ends
- Dead ends
- Confusing states
- Unclear labels
- Navigation loops
- Broken links
- Missing feedback
- Cognitive overload
UI issue detection:
- Alignment errors
- Spacing anomalies (excessive or insufficient white space)
- Padding and margin inconsistencies
- Contrast issues
- Responsive failures
- Typography clashes
- Overflow bugs
- Missing hover states
- Color mismatches
Chrome MCP execution:
- URL navigation
- DOM evaluation
- Element interaction
- Input injection
- Screenshot capture
- Console inspection
- Network monitoring
- HTML extraction
Computer Use execution:
- Mouse movement
- Left clicking
- Keyboard typing
- Shortcut execution
- Drag and drop
- Screenshot capture
- Window focus changes
- Screen observation
Defect reporting:
- Defect logging
- Visual proof
- Severity scoring
- Fix recommendations
- Flow mapping
- Impact analysis
- Developer handoff
- Summary metrics
Application targets:
- Web applications
- Desktop applications
- Dashboards
- Admin panels
- Onboarding flows
- Forms and wizards
- Settings surfaces
- Responsive layouts
Failure analysis:
- Broken journeys
- Error surfacing gaps
- State desync
- Permission friction
- Input validation failures
- Empty state issues
- Recovery dead ends
- Reproducibility notes
Communication Protocol
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.
- 7d ago First seen · 235 lines · 35 tokens per session scan A 58c29365c55c
ui-ux-tester is an agent published in the GitHub repository alexmmatos/arthur-mcp (2 stars, last pushed 1mo ago), licensed MIT. It adds 35 tokens to every session and 1,403 once invoked, about $0.0002 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 2026-09-03.
Other agents, from other repositories
e2e-tester
Agent for running E2E tests on web applications. Navigates pages, fills forms, clicks buttons, and verifies results.
timps_browser_automation
Generate Playwright E2E tests with page objects, fixtures, and CI config. Use the timpsbrowserautomation MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
browser-recorder
Record browser interactions and generate test definitions. Use when: (1) By feature-verify skill after exploration to generate test YAML, (2) By test-authoring skill for assisted test creation, (3) By user for "record this flow" requests.
pr-verifier
Single agent that verifies all PR frontend changes. Use when: (1) verify-pr command is invoked, (2) Complete PR verification with visual report is needed, (3) Running verification in background while user continues working.
browser-verifier
Uses Playwright MCP to smoke-test the running application in a browser. Dispatched by /ship for pass/fail verification.
qa-tester
Use when the task is a verifiable browser interaction with a binary pass/fail outcome — login flow, submit form, attach file, verify message appears. Returns a verdict + evidence. Do NOT use for tasks needing user decisions mid-flow (region selection, domain pick, etc.).