adversarial-ux-test

Roleplay a hostile user persona to find and triage UX friction in web applications.

Updated Sep 10, 2026
One-click install
npx skills add https://github.com/loteiron/ZeusAgent --skill adversarial-ux-test-loteiron
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: adversarial-ux-test
Source: https://github.com/loteiron/ZeusAgent/tree/main/optional-skills/dogfood/adversarial-ux-test
Command: npx skills add https://github.com/loteiron/ZeusAgent --skill adversarial-ux-test-loteiron

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Standard QA finds bugs but misses friction — a technically correct app can still be unusable for real humans. This Skill roleplays a worst-case, non-technical user to surface confusing terminology, excessive clicks, missing onboarding, accessibility issues, and cold-start problems, then filters the complaints to separate real UX bugs from persona noise. ## Core Features & Use Cases - Persona-Based Testing: Generate or adopt a specific hostile persona (age, tech comfort, core task) and browse the app fully in character, attempting their actual tasks rather than a feature tour. - In-Character Feedback Report: Write a visceral rant in the persona's voice covering the good, the bad, and showstoppers, with screenshots of every pain point. - Pragmatism Filter: A mandatory triage step that classifies each complaint as RED (real UX bug), YELLOW (low priority), WHITE (persona noise), or GREEN (feature request), creating tickets only for actionable items. - Use Case: Before a product launch, run an adversarial UX test on your staging site as a 58-year-old coach who only uses WhatsApp, then receive a filtered list of at most 10 tickets tagged "ux-review" with verbatim quotes and suggested fixes. ## Quick Start Run an adversarial UX test on my staging site as a grumpy non-technical user and create tickets for the real UX issues found.

Frequently Asked Questions about adversarial-ux-test

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run an adversarial UX test on my website?▼

Tell the agent to run an adversarial UX test on your URL, optionally specifying a persona type. The agent generates a hostile non-technical persona, browses the app in character attempting real tasks, then delivers a filtered report with tickets.

What is the pragmatism filter in persona-based UX testing?▼

The pragmatism filter is a mandatory triage step that classifies each persona complaint as RED (real UX bug), YELLOW (low priority), WHITE (persona noise), or GREEN (feature request). Only RED and GREEN items become tickets, preventing raw grumpy-user rants from shipping as work items.

Can I test a local development server with this approach?▼

No, the Skill requires testing on a staging or deployed app, not local dev. It also recommends registering as a new user rather than using pre-seeded admin accounts, since the cold-start experience reveals the most friction.

How many tickets does an adversarial UX test create?▼

A maximum of 10 tickets per session, focused on the worst issues. RED and GREEN items get individual actionable tickets with verbatim persona quotes, while YELLOW items are grouped into one catch-all ticket and WHITE items appear only in the report.

What if the persona finds no UX complaints?▼

Zero complaints means the persona is too tech-savvy, not that the product is perfect. Make the persona older, less patient, and more set in their ways. Conversely, zero WHITE (noise) items signals the product has genuine UX problems.