aidd-user-testing

March 22, 2026 · View on GitHub

Generates dual test scripts — human (think-aloud protocol) and AI agent (executable with screenshots) — from user journey specifications.

Why

Human testers catch usability friction that automated tests miss. AI agent tests provide repeatable, screenshot-documented coverage. Running both ensures journeys are validated from both perspectives.

Usage

Commands: /user-test <journey> (generate human and agent scripts), /run-test <script> (execute an agent script with screenshots).

Scripts are saved to plan/ — human scripts as *-human-test.md, agent scripts as *-agent-test.md. Journey data comes from plan/story-map/<journey-name>.yaml.

When to use

  • Creating user test scripts from journey specifications
  • Running automated user tests with screenshots
  • Validating user journeys from both human and AI perspectives