On point: Context Engineering is the discipline of systematically and at runtime filling the context window of language models with the right information in optimal form—far more comprehensive than prompt engineering.
Context Engineering is a technical discipline that describes how language models are supplied with the right information at the right time—far more comprehensive than the well-known prompt engineering. The term, coined by Andrej Karpathy, places the systematic assembly of the context window in focus.
While prompt engineering focused on formulating individual input sentences, Context Engineering goes further: it addresses the composition of the entire input document for language models. In modern production applications—autonomous agents, multi-stage workflows, RAG systems, code generators, or customer service bots—the input to the model is not a simple query, but a highly complex composed document that is constructed dynamically at runtime.
Context Engineering denotes the discipline of bringing the right content at the right time into the context window of a language model: in the right form, the right order, and the right scope. Anthropic describes it as an iterative process in which the curation phase occurs every time data is passed to the model—not as a one-time task, but as a continuous design decision.
The context window in production applications typically consists of multiple layers: a static system prompt for basic behavioral rules, the dynamic conversation history from previous interactions, and external knowledge through Retrieval-Augmented Generation (RAG), which provides documents, database entries, or web content at runtime. The central challenge lies in the fact that context windows are finite, and decisions about compression, selection, or discarding of information are not trivial—such as with the “Lost in the Middle” problem, in which information placed centrally is given less attention by the model.
For engineers, Context Engineering is relevant because it leads from pure prompt-tuning to systemic design: whoever masters the art and science of context composition builds more robust and effective AI systems that intelligently decide at runtime which data the model has available when and in what form.
Source: www.it-daily.net · Published 3 June 2026
Lumi AI News — AI-assisted curation in accordance with Art. 50 EU AI Act. Paraphrase and classification through Lumi News Pipeline v1.2.9.