Stop Prompting Like a Noob
Everyone's using ChatGPT to write tests. Most are doing it wrong.
This workshop takes you beyond “write test cases for login” and into the art of Context Engineering. This is where you teach the LLM what matters before you ask it what to do.
We’ll start simple: prompt like a pro. We do this until the prompt tricks stop working and context starts to matter. We then level up fast: inject SRS logic, layer bug taxonomies, pull CVEs on the fly, and wire it all up into a Langchain agent that remembers, reasons, and retrieves.
Expect more hands-on exercises than we’ll have time for, one meaty case study, and zero slide fatigue. By the end, you’ll stop treating an LLM like a vending machine and start using it like a context-driven test collaborator.
This isn’t AI hype. It’s applied thinking, powered by language models, grounded in test design.
Required
- Laptop with Remote Desktop Connection access (mstsc command on Windows / MS Remote Desktop app on Mac / Windows RDC clients on Linux). This access is needed for doing hands-on exercises on a remote Windows-based machine during the tutorial. It's better to try that your RDC access is working before coming to the tutorial.
- ChatGPT/Claude/Gemini account
- Curiosity (sposored by you) and Caffeine (sponsored by ATD).
