What Is Context Engineering? Explained For Software Testers | AI Tester Dictionary #029

Two testers give the same AI agent the same instruction: test the checkout flow. The first sends only that instruction, and gets back steps that could have been written by someone who never saw the product. The second also sends the requirements, the test data, the tools and the state of the application, and gets back a test run. Same model, same task, and the only difference was what the model had to work with. Context engineering means designing all the information a model receives before it acts. Six kinds of information can arrive: the instructions, the data it works on, the tools it can call, the current state of the application, the history of the session, and whatever it can look up. Most people only plan the first one, and the model cannot ask for anything you left out. A common mistake is to think that more context is always better. Good context is three things: the relevant information, in a structure the model can use, at the moment it needs it. If you send too little, the model fills the gaps with its own guesses. If you send too much, the important detail is buried on page forty. The episode also shows the context assembly step between the task and the model, the full context for one password-reset test, how the context grows on the second pass of a multi-step run, and four ways the context can be wrong: incomplete, outdated, conflicting and misleading. ⏱️ Chapters 0:00 Same task, two results 0:23 What the AI Tester Dictionary is 0:34 Six things arriving at once 1:01 A sentence, or an environment 1:25 The step in the middle 1:53 One job, assembled 2:28 Too little, too much 2:52 The second pass gets more 3:22 More is not better 3:40 Four ways it goes wrong 4:11 What to hold on to 🔑 Key takeaways • Context engineering is designing everything the model receives: instructions, data, tools, application state, session history and whatever it can look up. The instructions are only one part of six, and the model cannot ask for anything you left out • Prompt engineering writes the instructions; context engineering assembles everything around them. In a running system, that assembly is a real step between the task and the model, and anything missing from it does not exist as far as the model is concerned • Good context is the relevant information, in a structure the model can use, at the moment it needs it. Too little and the model guesses; too much and the important detail gets buried • Test what happens when the context is wrong: incomplete context gives invented details, outdated context gives false passes and failures, conflicting context gives different answers to the same input, and misleading context is where security problems come from 📚 In this series #001 — Artificial Intelligence #002 — Machine Learning #003 — Deep Learning #004 — Generative AI #005 — Large Language Models #006 — Transformer #007 — Diffusion Model #008 — Foundation Model #009 — Token #010 — Context Window #011 — Prompt #012 — System Prompt #013 — Prompt Engineering #014 — Temperature #015 — Top K and Top P #016 — Hallucination #017 — Embeddings #018 — Vector Database #019 — Semantic Search #020 — Retrieval-Augmented Generation #021 — Chunking #022 — Reranking #023 — Knowledge Graph #024 — Multimodal AI #025 — Function Calling #026 — Structured Output #027 — Grounding #028 — Fine-Tuning #029 — Context Engineering ← you are here ▶️ Previously #028 — Fine-Tuning: taking a model that already exists and training it further on your own examples.    • What Is Fine-Tuning? Explained For Softwar...   ▶️ Next in the series #030 — AI Agents: a model plus the system that lets it act on all of this. 🔔 Subscribe for practical AI and software testing content. #SoftwareTesting #AIForTesting #AITesterDictionary #ContextEngineering #PromptEngineering #LLM #QAEngineer #SDET #AITesting #TestAutomation #AIAgents #QualityEngineering #SoftwareQuality FOLLOW ME 📸 Web: https://softwaretestingtrends.com/ Discord:   / discord   LinkedIn:   / ambysan   Twitter:   / ambysan   Facebook:   / softwaretestingtrend   Medium:   / ambysan   BLOGGING 📝 Software Testing Trends: https://softwaretestingtrends.com/ TechJournal: https://ambreenkhan.dev/ ABOUT ME 👩‍💻 https://ambreenkhan.netlify.app/