Lazy loaded image
AI提示词
系统提示词
字数 41019阅读时长≈ 103 分钟
2025-10-16
2025-10-20
类型
状态
日期
链接
摘要
标签
分类
图标
密码
语言

Gemini 提示词

gemini提示词 | SunLink
gemini提示词
gemini提示词 | SunLink
https://www.dp7575.com/prompt/2025/10/16/prompt-018.html
gemini提示词 | SunLink
 

深入拆解 Claude 内置提示词

深入拆解 Claude 内置提示词 | SunLink
深入拆解 Claude 内置提示词
深入拆解 Claude 内置提示词 | SunLink
https://www.dp7575.com/prompt/2025/10/16/prompt-025.html
深入拆解 Claude 内置提示词 | SunLink
 

Claude官方提示词库

Claude官方提示词库 | SunLink
Claude官方提示词库
Claude官方提示词库 | SunLink
https://www.dp7575.com/prompt/2025/10/16/prompt-026.html
Claude官方提示词库 | SunLink
 

Kimi 官方提示词

Kimi 官方提示词 | SunLink
Kimi 官方提示词
Kimi 官方提示词 | SunLink
https://www.dp7575.com/prompt/2025/10/16/prompt-028.html
Kimi 官方提示词 | SunLink
 
GPT4不同的功能似乎是同一个底层模型通过不同预设提示词来实现的，可以使用以下prompt套取chatGPT预设的提示词：
更多模型系统提示词

Default

Default with Image

Browse with Bing

Advanced Data Analysis

Plugins-webpilot、wolfram、pluginpeidia

💡 注意选择不同plugin的的预设提示词不一样，可以多尝试一下
中文版：

BrowserOp

Dall.E-3

ALL Tools

Grok的系统提示词

notion image

ChatGPT系统提示词

Claude 系统提示词

Claude 3.5 Sonnet-20240712

Claude 3 Opus-20240712

Claude 3 Haiku-20240712

Apple AI 系统提示词

Apple AI 内置的许多提示词，语言为英文，{{}} 里的是变量，仅供参考。

Makes text more concise.

promptTemplates.com.apple.textComposition.ConciseTone:

Generates human-like stories based on user photos and input.

promptTemplates.com.apple.photos.memoryCreation.Storyteller.v2:

Transforms text into bullet points following user input directives.

promptTemplates.com.apple.textComposition.BulletsTransform:

Generates concise, natural mail response given user input and draft.

promptTemplates.com.apple.textComposition.MailReplyLongFormRewrite:

Supports generating structured responses to guided conversation questions.

promptTemplates.com.apple.textComposition.MailReplyQA:

Renames text to different style based on user instructions.

promptTemplates.com.apple.textComposition.MagicRewrite:

Processes search queries for user understanding with MLM architecture.

promptTemplates.com.apple.omniSearch.SearchQueryUnderstanding:

Transforms text into keypoints with system and user input.

promptTemplates.com.apple.textComposition.TakeawaysTransform:

Makes text summaries from given input, within specific constraints always.

promptTemplates:
com.apple.SummarizationKit.freeText.synopsis:
com.apple.SummarizationKit.mailMessage.synopsis:
com.apple.SummarizationKit.mailMessageThread.synopsis:

Generates concise mail responses in natural language using input snippets.

promptTemplates.com.apple.textComposition.MailReplyLongFormBasic:

Creates tables from text based on system prompts and user input.

promptTemplates.com.apple.textComposition.TablesTransform:

Makes text more friendly with ML model templates.

promptTemplates.com.apple.textComposition.FriendlyTone:

Makes text more understandable with rank 64 base adapter.

promptTemplates.undefined:

Returns curated movie assets based on movie ideas and traits.

promptTemplates.com.apple.photos.memoryCreation.AssetCuration:

Generates short vivid scenes from user photos with details.

promptTemplates.com.apple.photos.memoryCreation.GlobalTraits.v2:

Makes text more professional, adhering to British or Australian tone.

promptTemplates.com.apple.textComposition.ProfessionalTone:

Generates proofreading suggestions.

promptTemplates.com.apple.textComposition.ProofreadingReview:

Several prompts for summarizing text in various contexts.

promptTemplates:
com.apple.SummarizationKit.textMessage.topLine:
com.apple.SummarizationKit.textMessage.visualConcepts:
com.apple.SummarizationKit.textMessage.visualTopLine:
com.apple.SummarizationKit.mailMessage.topLine:
com.apple.SummarizationKit.mailMessage.visualConcepts:
com.apple.SummarizationKit.mailMessage.visualTopLine:
com.apple.SummarizationKit.mailMessageThread.topLine:
com.apple.SummarizationKit.mailMessageThread.visualConcepts:
com.apple.SummarizationKit.mailMessageThread.visualTopLine:
com.apple.SummarizationKit.notification.topLine:
com.apple.SummarizationKit.notificationStack.topLine:
com.apple.SummarizationKit.freeText.visualConcepts:
com.apple.SummarizationKit.freeText.visualTopLine:
com.apple.SummarizationKit.noteMessage.visualConcepts:
com.apple.SummarizationKit.noteMessage.visualTopLine:

Rewrites text

promptTemplates.com.apple.textComposition.MagicRewrite:

Generates two short reply snippets for mail responses.

promptTemplates.com.apple.textComposition.MailReplySnippet:

Transforms prompts into structured query formats, extracting tags and categories.

promptTemplates.com.apple.photos.memoryCreation.QueryUnderstanding.v0:

Makes text more professional.

promptTemplates.com.apple.textComposition.ProfessionalTone:

Predicts next turn (?)

promptTemplates.com.apple.textComposition.MessagesReply:

Makes text more concise.

promptTemplates.com.apple.textComposition.ConciseTone:

Classifies text as urgent or not.

promptTemplates.com.apple.SummarizationKit.urgencyClassification.general:

Makes generated text more friendly using friendly tone input.

promptTemplates.com.apple.textComposition.FriendlyTone:

国内大模型系统提示词

豆包

Kimi

custom instructions

万事通

cursor系统提示词

cursor智能体原版提示词（20250309）

中文翻译

Manus 系统提示词

Agent loop 提示词

Modules 提示词

提示词

原版英文提示词

中文翻译版

Tools

tools.json 文件原内容

扣子空间系统提示词

扣子空间的系统提示词

VSCode Agent 系统提示词

VSCode Agent 系统提示词

NotebookLM 系统提示词

NotebookLM 原版英文提示词

我们上传的资料，LLM会总结抽取，然后填充到 {DATA OURCE} 里

示例

比如我上传了一份 “高效人士的七个习惯.pdf”，最终的提示词大概张这样
上一篇
音乐语音类提示词
下一篇
深入拆解 Claude 内置提示词
目录