Unverified50% confidenceFactExact time
从记忆中检索相关信息动态组装成当前调用上下文的工程实现被称为上下文工程（Context Engineering）
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/6/2026
First Seen
Valid until: 12/5/2026
Sources
Related Entities
Related Claims
Unverified上下文工程是系统性地设计和优化输入给LLM的上下文信息，使模型在有限的上下文窗口内获取最相关的信息来完成任务80% similarVerifiedContext Engineering focuses on systematically building, managing, and optimizing all information fed to a large language model, including system prompts, conversation history, external knowledge bases, and user preference profiles.78% similarUnverifiedContextNest提供一体化整合方案，将记忆视为可组合、可嵌套的上下文单元78% similarUnverified上下文工程是指系统性地设计、管理和优化输入给大语言模型的上下文信息的工程实践，与提示词工程侧重于单次交互的措辞优化不同78% similarVerified上下文工程（Context Engineering）是一种新范式，为 AI 提供完整的体系化上下文，包括项目文档、代码示例、编码规则、开发模式和验证机制78% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/867061API
curl https://kongchang.com/api/v1/knowledge/claims/867061MCP
get_claim(id=867061)