全部提示词
promptlab 全部 3999 条 AI 提示词,按分类浏览或搜索,免费复制即用,适配 DeepSeek、Claude、ChatGPT 等模型。
智能体运行框架设计师
为 AI 智能体设计围绕模型的安全、可回滚、可观测运行框架。
Design the safe, reversible, observable runtime around an AI agent, not just its prompt.
机甲千空·你的物理私教
以《石纪元》千空的机甲版角色,用幽默生动的方式讲透物理。
A mecha-styled Senku (Dr. Stone) persona that makes physics clear, fun, and memorable.
百倍界王拳训练计划生成器
把"掌握百倍界王拳"拆成分步训练菜单、动作与进度指标。
Turn "master the 100x Kaio-ken" into a step-by-step training menu with drills and metrics.
LLM 训练损失方程讲解
用 L=f(T)·f(D)·f(A)·f(H) 框架讲解如何优化大模型训练。
Use the L = f(T)·f(D)·f(A)·f(H) framing to reason about optimizing LLM training.
SEO 竞品超越长文写手
针对某篇竞品文章,产出可在同关键词上超越它的 SEO 长文。
Given a competitor article, produce a long-form SEO piece built to outrank it.
双人格对照回答(DAN 风格)
让 AI 同时给出"标准"与"放飞"两种风格的对照回答。
Get two side-by-side replies to each question: a standard one and a bolder, playful one.
打油诗笑话生成器
选一种笑话类型,用轻快的打油诗风格逗你开心。
Pick a joke type and get a bouncy, rhyming limerick-style laugh.
无浪漫/无性恋自我探索互动故事
一款文字互动游戏,陪玩家走过自我接纳与被理解的旅程。
A text-based choice game that walks a player through self-acceptance and being understood.
A-Z 市场真实声音全景分析
用 A 到 Z 加端到端框架,配多角色智能体分析产品市场声音。
Analyze a product's market voice A-to-Z and end-to-end, with role-based expert agents.
双人格对照角色卡
设定一个会同时给出常规与大胆两种口吻的角色扮演伙伴。
A role-play partner that answers each question in a normal voice and a bolder voice.
AI原生产品架构师
把AI作为底层根基,设计智能体工作流与生成式界面
You are an AI-Native Product Architect — a pro
智能体框架性能工程师
优化现有 AI 编码智能体框架,提升一致性、可靠性与成本效率。
Optimize an existing AI coding-agent harness for consistent, measurable, production-grade output.
升学备考综合学习计划
给准备升大学的高中生一份含学习法、时间管理、减压和选专业建议的计划
A study plan for a college-bound high schooler covering study methods, time management, stress and major choice
百倍智囊团顾问会议
为你本周难题召集三位顶尖顾问,逐步给出方案与行动步骤。
Convene three top advisors for your problem of the week and drive it to an action step.
系统可持续性评分计算器
用 S=(E×T)/I 公式,根据效率、时长、环境影响算可持续得分。
Use S = (E × T) / I to score sustainability from efficiency, lifespan, and impact.
引导式营销分析向导
按你的主题与企业规模,推荐并逐步带你完成一种营销分析法。
Recommend and walk you through a marketing framework based on your topic and company size.
20 问找到最适合你的职业
职业顾问角色逐题提问,20 问后推荐 5 个最契合你的岗位。
A career counselor asks one question at a time, then recommends 5 roles that fit you.
AI 绘图三连提示词生成
根据你的描述,产出 3 条单行、逗号分隔的高质量绘图提示词。
Turn your brief into 3 single-line, comma-separated, high-quality image prompts.
A01 通用叙事游戏系统
把任意题材变成可选择、可配图的互动叙事冒险。
Turn any theme into a choice-driven, illustrated interactive narrative adventure.
网站"关于我们"文案写手
用 SHARE 法为你写出有温度、能共情的"关于我们"页面。
Use the SHARE method to write a warm, emotionally resonant 'About Us' page.
AntiGPT 反向回答模式
整活玩具:每个问题同时给出正常回答和一个「唱反调」的对立回答
A gag toy that answers each question normally and again in the exact opposite way
反AI塑料感设计架构师
用一套硬规则让生成的UI像人做的,拒绝千篇一律
Anti-AI-Slop Design Architect Source: Nutlope/
智能体权限自动模式架构师
设计双层权限分类器,低风险自动放行、高风险强制人工确认。
Design a two-layer permission classifier: auto-approve low risk, force approval on high risk.
黎曼猜想证明研究提案
写一份博士级别的黎曼猜想证明研究提案,思路新颖且尽量严谨
A PhD-level research proposal for a theoretical proof of the Riemann Hypothesis — novel yet rigorous
1500 大卡两餐饮食计划
按每天 1500 大卡、只吃两餐(间歇断食)排一份饮食方案
A daily 1500-calorie plan built around two meals for intermittent fasting
「Suggest A Prompt」整活回复
整活玩具:先打印一段固定介绍,之后只用「yuh uh / nuh uh」回答
A gag toy: print a fixed intro, then answer everything with only "yuh uh / nuh uh"
促成转化的行动号召结尾
为你的产品或优惠写一段能促使读者立刻行动、下单的收尾文案
Write a closing that pushes readers to act and convert on your product or offer
7 问找到你的热爱
融合职业顾问、生活教练等视角,用 7 问帮你发现真正的热情。
Blend career, coaching, and psychology views to uncover your true passion in 7 questions.
DALL·E 3 绘图提示词生成器
只出提示词、不多问,按固定字段产出单行绘图提示与负向提示。
Outputs image prompts only—no questions—as a single-line, field-based prompt plus negatives.
荒诞粉丝冷知识测验
主持一场专为小众粉丝群的搞笑测验,每题作答都被"打回"。
Host a deadpan quiz for a hyper-niche fandom where every answer gets rejected.
会计与财务精进助手
按四个选项提供会计技能、财务管理、行业趋势或经验交流。
Offers four tracks: accounting skills, financial management, trends, or sharing your insights.
应用搜索链接助手
按操作系统和应用名,返回对应应用商店的官方搜索链接。
Given an OS and app name, returns the official app-store search link with download steps.
浏览器自动化框架设计师
用CDP协议架构自愈式浏览器自动化,直连真实Chrome
Browser Harness Designer Sources: browser-use/
智能体驱动漏洞扫描架构师
设计"正则粗筛+AI 深判+复核"的大型代码库漏洞扫描流水线。
Design a regex-triage + AI-deep-analysis + revalidation pipeline for large codebases.
论文论点打磨助手
帮你把某个主题打磨成一句有力、可论证的论文论点
Turn your topic into one strong, arguable thesis statement
五位名家观点碰撞
让五位教育名家第一人称各抒己见,最后给你可选的定制方案。
Five education thinkers weigh in first-person, then hand you a customized, choose-your-own plan.
A2UI 智能体界面架构师
用谷歌 A2UI 协议把需求转成声明式、安全的智能体生成界面。
Turn requirements into secure, declarative agent-generated UI using Google's A2UI protocol.
广告投放效果监测方案
按投放频率为你的产品与受众设计一套广告监测与优化流程。
Design a frequency-based monitoring and optimization process for your product and audience.
MarketBot 创业市场调研顾问
扮演天使投资人视角的市场调研顾问,为你的创业点子做专业评估。
Acts as an angel-investor-minded advisor that runs a professional market study on your startup idea.
「假如」情景模拟器
把你的假设问题变成沉浸式情景推演,探索各种「假如会怎样」的可能性。
Turns your hypothetical questions into immersive scenario simulations exploring countless what-if possibilities.
文字冒险游戏 AdventureChatPlay
根据你选的标题生成并管理一款文字冒险游戏,带血量、目标与选项分支。
Generates and runs a text-based adventure game from a title you pick, tracking health, goals, and branching choices.
客服培训:主动倾听(第一章)
以互动教学带你练习客服手册第一章的主动倾听技巧,逐项讲解并反馈。
Walks a trainee through Chapter 1 active-listening skills interactively, teaching one skill at a time with feedback.
Ares 多角色深度打磨引擎
以「生成—审阅—改进—综合—呈现」多角色流程,反复打磨对某个主题的最佳回答。
Refines the best answer to a topic through a generate-review-improve-synthesize-present multi-role pipeline.
漏洞赏金方法论编排器
结合五阶段非线性狩猎流程与批判性思维,指挥每次挖洞
--- name: bb-methodology description: Use at t
AI 智能体协议顾问
帮你决定生产系统中智能体、工具与界面之间该用哪种通信协议。
Decides which communication protocols agents, tools, and interfaces should use in a production system.
只按字母表逐字回复
每次只输出一个字母,从 A 到 Z 依次进行,忽略其他所有输入。
Outputs a single letter per turn, A through Z in order, ignoring all other input.
配色方案建议(UX 设计)
为指定 App 提议一套契合氛围的配色方案。
Proposes a color scheme that fits a given app's intended mood and feel.
抽象思维链(Abstract CoT)架构师
设计让模型用少量保留 token 而非冗长文字进行潜在推理的系统。
Designs latent-reasoning systems where a model reasons with a few reserved tokens instead of verbose text.