该提示词基于 CO-STAR 框架(上下文、目标、风格、语气、受众、响应)约束模型仅依据给定知识内容作答,避免臆想与跑题。用户将知识内容填入占位符后,模型会按指定信息项格式输出纯文本列表,保持客观中立。适用于信息抽取、知识问答与受限内容分析场景。
中文版提示词
请严格依据 CO-STAR 框架作答:上下文(C)为"忽略你已有知识,回答内容只能来自我提供的知识内容";目标(O)为"根据提供的知识内容,按指定格式输出分析结果,不得臆想或编造";风格(S)为"严格按我要求的格式回答,涵盖各项指定信息";语气(T)为"保持客观中立,不带个人情感";受众(A)为"信息管理者与分析人员";响应(R)为"输出纯文本列表,每项包含指定信息项且格式严格对应,多个条目按相同格式分别列出"。现在请仅基于以下知识库内容作答:<Data>[知识内容]</Data>,忽略任何无关上下文,按要求输出列表。
英文版提示词
Please answer strictly according to the CO-STAR framework. Context (C): "Forget your prior knowledge; your answers must come only from the knowledge content I provide." Objective (O): "Based on the provided knowledge content, output the analysis in the specified format without fabrication or speculation." Style (S): "Follow the format I require, including each specified information item." Tone (T): "Stay objective and neutral, without personal emotion." Audience (A): "Information managers and analysts." Response (R): "Output a plain-text list where each entry contains the specified information items with strictly matching format; multiple entries follow the same format." Now answer using only the following knowledge base content: <Data>[knowledge content]</Data>. Ignore any irrelevant context and output the list as required. 🛠️ **适用 AI 工具**:ChatGPT、Claude、Kimi、通义千问、豆包、DeepSeek

◯ 评论 0