Done isn't a test result. How your AI agent should hand off UI work

Описание и применение

Агенты теперь предоставляют прозрачные заметки о сдаче, указывая, что было протестировано и что осталось непроверенным, что позволяет целенаправленно запрашивать недостающие проверки.

Задача

AI agents often claim completion without evidence, leading to unverified UI changes.

Результат

Agents now produce transparent handoff notes that specify what was tested and what remains unverified, enabling targeted follow‑up checks.

Рабочий процесс

• 1. Define acceptance criteria and observable acceptance criteria.
• 2. Agent delivers artifact, decisions, and assumptions.
• 3. Agent labels each result with evidence level (Proposed, Heuristic review, Automated check, Manual test, User research).
• 4. Agent lists untested items.
• 5. Reviewer reviews handoff note and requests missing checks if needed.

Доказательства

Design Decision Method's last step is there to stop that. Step 6, deliver and validate, asks the agent to hand over the work, the decisions behind it, and a short validation status that says what each result rests on.

Источник

Each label says what a result can support and what it can't: ...

Источник

Every line says what it rests on. You can see that mobile keyboard behaviour is the next check, and that nobody has tried the form with real users.

Источник