An assistant that completes your code is fine. An agent that reads your repo, runs things, and fixes itself is a different story. How I use it in QA.
- AI
- Claude
- Agents
- Productivity
- QA
Notes on QA, automated testing, and AI — from the trenches.
An assistant that completes your code is fine. An agent that reads your repo, runs things, and fixes itself is a different story. How I use it in QA.
I've been using AI to automate E2E tests for a while now. It saves me hours, but not for the reasons the Twitter threads claim.