Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
- Added SKILL.md metadata specifying required config file (~/.openclaw/config/nova-act.json) and Python3 binary. - No code or feature changes; documentation and metadata update only.
What This Skill Does
Runs browser-based usability tests on websites using Amazon Nova Act, simulating multiple user personas with different goals and tech proficiency levels. The agent generates personas, executes tests step-by-step, interprets raw results, and compiles findings into an HTML report. Tests stop automatically before actions with real-world impact like purchases, account creation, or public posts.
Automates multi-persona usability testing with real browser sessions, removing the need to recruit manual testers or configure separate testing services.
When to use it
- Testing an e-commerce checkout flow before launch
- Checking if elderly users can find contact info on a healthcare site
- Auditing a SaaS onboarding experience for first-time users
- Evaluating a hotel booking flow without completing a reservation
- Identifying navigation friction on a government or nonprofit website
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Test the checkout flow on myshop.com for a first-time buyer
- 1Generates a first-time buyer persona with age, tech proficiency level, and specific goals
- 2Runs the Nova Act test script against myshop.com, simulating product search and cart steps
- 3Stops automatically before payment confirmation to avoid real transactions
- 4Reads raw JSON results and interprets whether each step's goal was achieved
- 5Generates an HTML report with friction points, step outcomes, and browser trace links
HTML usability report saved to nova_act_usability_report.html with per-persona results and friction analysis
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.