Context engineering supersedes prompt engineering¶
Insight: "Context engineering" — constructing the entire information environment an AI needs to solve a problem — has emerged as the successor to "prompt engineering." Popularized by Shopify CEO Tobi Lütke and Andrej Karpathy in mid-2025, the term reflects that real-world LLM applications succeed not through cleverly worded prompts but through carefully assembling the full context window: relevant code, design docs, error logs, schemas, examples, and constraints. "Prompt engineering was about cleverly phrasing a question; context engineering is about constructing an entire information environment."
Detail: Practical context engineering means: be precise (vague requests → vague answers), provide relevant code files, include design documents, share full error logs, show database schemas, use PR feedback as context, give examples of desired output, and clearly state constraints. The shift is structural: prompt engineering ends once you craft a good prompt, whereas context engineering begins with designing whole systems that bring in memory, knowledge, tools, and orchestration. This aligns with Cline's move from RAG to "agentic search" and the broader pattern of tools like CLAUDE.md, rule files, and memory systems.
Sources
Related: existing entry "Prompt engineering for coding is a systematic skill" in external/claude-code.md — SUPERSEDES