Cursor found that sections of their prompt that had been effective with earlier models needed tuning ... <maximize_context_understanding> Be THOROUGH ... they found it counterproductive with GPT-5
Generally speaking, reasoning models will provide better results on tasks with only high-level guidance. This differs from GPT models, which benefit from very precise instructions.
Wrap examples in `<example>` tags (multiple examples in `<examples>` tags) so Claude can distinguish them from instructions. ... XML tags help Claude parse complex prompts unambiguously
06术语
术语标记敏感度的迁移性 Transferability of Markup Sensitivity
界定对某种标记(如 XML 标签)的响应强度取决于该模型是否针对此类标记做过指令对齐训练,因此同一标记在不同模型上的效果不可假定一致。
相关节点—
术语分隔符与上下文隔离 Delimiters and Context Isolation
界定用 XML 标签、专用分节标记等把指令、示例、资料与用户输入在文本上划出边界,使模型能无歧义解析多部分提示。
相关节点—
术语换模型即失效与重编译 Model Swap Invalidates and Requires Recompilation