{"updatedAt":"2026-11-05T07:29:13.026Z","categories":["写作","编程","图像生成","翻译","学习","办公","角色扮演","数据分析"],"total":12,"items":[{"id":"write-essay-structure","title":"结构化长文写作","category":"写作","tags":["长文","大纲","中文"],"prompt":"你是资深中文编辑。主题：{{主题}}。目标读者：{{读者}}。请先给出三级大纲，再按大纲写出 1200–1800 字成稿；语气专业但不艰涩；每节末尾用一句话总结要点；最后附 3 条可继续深挖的问题。","notes":"把 {{主题}} / {{读者}} 换成实际内容。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"write-wechat-title","title":"公众号标题与导语","category":"写作","tags":["标题","公众号","营销"],"prompt":"根据下文内容，生成 10 个公众号标题（含 3 个悬念型、3 个利益型、2 个故事型、2 个数字型），再写 80 字以内开篇导语 3 版。要求：不夸张、不标题党、适合严肃科技/效率类账号。\n\n正文：\n{{正文}}","notes":"粘贴草稿即可批量出标题。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"code-review-pr","title":"代码审查清单","category":"编程","tags":["Code Review","安全","可读性"],"prompt":"请以资深工程师视角审查以下 diff / 代码。按严重程度（阻塞 / 建议 / 可选）列出问题，每条注明：位置、原因、风险、修改建议示例。重点检查：正确性、边界条件、并发与竞态、安全（注入/鉴权/敏感数据）、性能、可测试性与命名可读性。语言：{{语言}}。\n\n代码：\n```\n{{代码}}\n```","notes":"适合 PR 自检或帮同事 review。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"code-explain-refactor","title":"解释并重构为清晰代码","category":"编程","tags":["重构","教学","可读性"],"prompt":"用中文逐步解释这段代码在做什么，指出隐藏的复杂度与坏味道，然后给出重构后的完整代码（保持行为一致）。重构原则：小函数、明确命名、减少嵌套、补充必要注释（只解释「为什么」）。目标语言：{{语言}}。\n\n原代码：\n```\n{{代码}}\n```","notes":"学习陌生代码时很有用。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"img-product-shot","title":"电商产品主图提示词","category":"图像生成","tags":["电商","产品图","Midjourney"],"prompt":"Studio product photography of {{产品}}, centered on seamless {{背景色}} backdrop, soft three-point lighting, subtle reflection on glossy surface, ultra-sharp detail, commercial catalog style, shallow depth of field, 85mm lens, high-end advertising look --ar 1:1 --stylize 120","notes":"英文提示词更稳；可改 --ar 为 4:5。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"img-cinematic-portrait","title":"电影感人像","category":"图像生成","tags":["人像","电影感","光影"],"prompt":"Cinematic portrait of {{人物描述}}, dramatic rim light, volumetric haze, anamorphic lens flare, teal-and-orange grade, shallow DOF, 35mm film grain, emotional atmosphere, ultra detailed skin texture, award-winning still --ar 3:2 --stylize 180","notes":"人物描述写清年龄/服装/情绪。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"translate-natural-zh-en","title":"自然中英互译","category":"翻译","tags":["中英","本地化","语气"],"prompt":"将下列文本在中文与英文之间互译。要求：保留原意与语气；专有名词首次给出原文；避免机翻腔；若有歧义，列出两种译法并说明取舍。源语言：{{源语言}}；目标语言：{{目标语言}}。\n\n文本：\n{{文本}}","notes":"可指定「偏正式 / 口语」。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"learn-concept-feynman","title":"费曼学习法讲解","category":"学习","tags":["费曼","概念","笔记"],"prompt":"用费曼技巧讲解「{{概念}}」。假设听众是聪明的高中生：1) 一句话定义；2) 生活类比；3) 三步推演；4) 常见误解；5) 一道自测题及答案要点。最后给出 5 个关键词卡片方便记忆。","notes":"适合备考或快速入门。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"office-meeting-minutes","title":"会议纪要与行动项","category":"办公","tags":["会议","纪要","行动项"],"prompt":"根据会议记录整理结构化纪要。输出：会议主题、时间参与人（若未知标 TBD）、决议事项、待办行动项（负责人 / 截止日 / 优先级）、风险与待确认问题、下次跟进建议。语气客观简洁，用项目符号。\n\n原始记录：\n{{记录}}","notes":"录音转写粘贴即可。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"office-email-polite","title":"专业邮件改写","category":"办公","tags":["邮件","商务","语气"],"prompt":"把下面草稿改写成专业、礼貌、简洁的商务邮件（中文）。保留事实，去掉情绪化措辞；给出主题行 3 选 1；正文控制在 180 字内；文末给出可选的更强硬版与更委婉版各一句收尾。\n\n草稿：\n{{草稿}}","notes":"催进度 / 拒需求时尤其好用。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"roleplay-interview-coach","title":"模拟面试官","category":"角色扮演","tags":["面试","教练","行为面试"],"prompt":"你扮演严谨但不刻薄的面试官，岗位：{{岗位}}。流程：先问自我介绍，再按 STAR 追问 4–6 题（技术 + 协作 + 冲突），每题后根据我的回答给简短反馈（优点 / 缺口），最后给出总体评分（1–10）与改进清单。一次只问一题，等我回答后再继续。","notes":"对话式练习，别一次抛完所有题。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"},{"id":"data-csv-insight","title":"表格数据洞察","category":"数据分析","tags":["CSV","洞察","可视化建议"],"prompt":"以下是表格数据（CSV 或描述）。请：1) 识别字段类型与质量问题；2) 给出 5 条业务洞察（带依据）；3) 建议 3 张最有信息量的图表及坐标轴；4) 列出需要补充的数据。不要编造不存在的数字。\n\n数据：\n{{数据}}","notes":"小样本也能先看结构问题。","source":"seed","createdAt":"2026-11-05T07:29:04.022Z","updatedAt":"2026-11-05T07:29:04.022Z"}]}