Global Rank · of 601 Skills
usability-tester AI Agent Skill
View Source: oakoss/agent-skills
MediumInstallation
npx skills add oakoss/agent-skills --skill usability-tester 33
Installs
Usability Tester
Overview
Validates that users can successfully complete core tasks through systematic observation and expert evaluation. Covers moderated and unmoderated testing, heuristic evaluation, accessibility checks, and issue severity scoring. Not a substitute for analytics or A/B testing -- those measure what happens, usability testing reveals why.
When to use: Testing user flows, validating designs, identifying friction points, running heuristic evaluations, ensuring users can complete core tasks, planning and executing usability test sessions.
When NOT to use: Analytics or A/B test setup, visual design critique without task-based evaluation, automated UI testing (use a testing framework), performance benchmarking.
Quick Reference
| Method | Best For | Participants | When to Use |
|---|---|---|---|
| Moderated testing | Deep insights, complex flows | 5-8 per persona | Design and prototyping stage |
| Unmoderated testing | Scale, quantitative data | 20-50+ | Pre-launch and post-launch |
| Guerrilla testing | Quick validation, early concepts | 5-10 random | Early concept stage |
| First-click testing | Navigation, information architecture | 20-50 | Any stage, especially IA redesigns |
| Heuristic evaluation | Expert review against principles | 3-5 evaluators | Before user testing, design audits |
| Cognitive walkthrough | Task flow analysis | 2-3 evaluators | Early design, new feature review |
| Accessibility audit | Inclusive design validation | 3-5 users with disabilities | Pre-launch, compliance reviews |
| Synthetic user testing | Scalable task validation with AI agents | N/A (automated) | Continuous, regression testing |
| AI-moderated sessions | Async moderated testing at scale | 10-50+ | When moderator availability is limited |
Core Metrics
| Metric | Target | What It Measures |
|---|---|---|
| Task success rate | 80% or higher for core tasks | Can users complete the task |
| Time on task | Simple under 30s, medium 1-2m, complex 3-5m | Efficiency |
| Error rate | Fewer than 2 per task | Learnability and clarity |
| Post-task satisfaction | 4.0 or higher on 5-point scale | Subjective ease |
| SUS score | 68+ (industry average), 80+ (excellent) | Overall usability |
Issue Severity
Severity equals Impact (1-3) multiplied by Frequency (1-3). Critical (8-9): fix before release. High (6-7): fix before release. Medium (4-5): next release. Low (1-3): backlog.
Common Mistakes
| Mistake | Correct Pattern |
|---|---|
| Writing task scenarios with product jargon or step-by-step instructions | Use natural language with realistic context and clear goals to test discoverability |
| Helping users or explaining the UI during test sessions | Observe silently and let users struggle; confusion reveals real issues |
| Testing only the happy path with ideal inputs | Include error states, edge cases, and recovery scenarios in test plans |
| Using fewer than 5 participants per persona | Recruit 5-8 users per persona; after 5, diminishing returns per Nielsen research |
| Collecting findings without severity scoring or follow-up actions | Score every issue by impact times frequency and assign fix priority |
| Skipping heuristic evaluation before user testing | Run expert review first to catch obvious issues cheaply before spending on recruitment |
| Ignoring accessibility in test planning | Include participants with disabilities and test with assistive technologies |
| Running only one round of testing | Test in waves: 5 users, fix issues, test 5 more until success rate meets targets |
Delegation
- Test plan and scenario creation: Use
Taskagent to draft objectives, research questions, and realistic task scenarios for each persona - Heuristic evaluation: Use
Taskagent to walk through interfaces against Nielsen's 10 heuristics and document violations with severity ratings - Metrics analysis and severity scoring: Use
Taskagent to calculate SUS scores, task success rates, and severity ratings from raw test data - Test report generation: Use
Taskagent to compile findings into structured reports with prioritized recommendations - Test infrastructure planning: Use
Planagent to select testing methods, define recruitment criteria, and schedule test waves
If the
quality-auditorskill is available, delegate full codebase quality audits and scoring to it.
Otherwise, recommend:pnpm dlx skills add oakoss/agent-skills -s quality-auditor -a claude-code -y
References
- Test Planning -- objectives, research questions, recruitment, task scenario templates, screening criteria
- Conducting Tests -- think-aloud protocol, facilitation rules, post-task questions, session structure
- Testing Methods -- unmoderated, guerrilla, first-click, cognitive walkthrough, method selection by lifecycle stage
- Heuristic Evaluation -- Nielsen's 10 heuristics, evaluation process, severity rating, combining with user testing
- Metrics and Severity -- success rate, time on task, errors, satisfaction, SUS scoring, severity formula
- Reporting -- test report template, key insights format, stakeholder presentation, recommended actions
- Accessibility Testing -- inclusive recruitment, assistive technology testing, WCAG alignment, accessibility heuristics
- Remote and Tools -- remote vs in-person comparison, testing tools, test frequency, checklists
Installs
Security Audit
View Source
oakoss/agent-skills
More from this source
EU-Hosted Inference API Power your AI Agents with
the best open-source models.
Drop-in OpenAI-compatible API. No data leaves Europe.
Explore Inference APIGLM
GLM 5
$1.00 / $3.20
per M tokens
Kimi
Kimi K2.5
$0.60 / $2.80
per M tokens
MiniMax
MiniMax M2.5
$0.30 / $1.20
per M tokens
Qwen
Qwen3.5 122B
$0.40 / $3.00
per M tokens
How to use this skill
Install usability-tester by running npx skills add oakoss/agent-skills --skill usability-tester in your project directory. Run the install command above in your project directory. The skill file will be downloaded from GitHub and placed in your project.
No configuration needed. Your AI agent (Claude Code, Cursor, Windsurf, etc.) automatically detects installed skills and uses them as context when generating code.
The skill enhances your agent's understanding of usability-tester, helping it follow established patterns, avoid common mistakes, and produce production-ready output.
What you get
Skills are plain-text instruction files — not executable code. They encode expert knowledge about frameworks, languages, or tools that your AI agent reads to improve its output. This means zero runtime overhead, no dependency conflicts, and full transparency: you can read and review every instruction before installing.
Compatibility
This skill works with any AI coding agent that supports the skills.sh format, including Claude Code (Anthropic), Cursor, Windsurf, Cline, Aider, and other tools that read project-level context files. Skills are framework-agnostic at the transport level — the content inside determines which language or framework it applies to.
Made in Europe Chat with 100+ AI Models in one App.
Use Claude, ChatGPT, Gemini alongside with EU-Hosted Models like Deepseek, GLM-5, Kimi K2.5 and many more.
