repurpose-instagram

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Instagram Content Generator

Instagram 内容生成器

Generate carousel script, caption, and reel script from content atoms.
根据 Content atoms 生成轮播脚本、Caption 和 Reel 脚本。

Inputs

输入

You receive from the orchestrator:
  • Content atoms (typed and impact-rated)
  • Main argument, target audience, primary topic
  • Brand voice profile (5-dimension scores or override flag)
  • Flags:
    --brief
    (caption only),
    --images
    (generate image prompts)
你从 orchestrator 收到:
  • Content atoms(已完成类型标注和影响力评级)
  • 核心论点、目标受众、主要主题
  • 品牌语气档案(5 维评分或 override 标志)
  • Flags:
    --brief
    (仅生成 Caption)、
    --images
    (生成图像提示词)

References

参考资料

Load before generating:
  • references/platform-specs.md
    -- carousel specs, reel dimensions, algorithm weights
  • references/hook-formulas.md
    -- first-slide hooks, caption openers, reel hooks
  • references/voice-adaptation.md
    -- Instagram tone rules (relaxed, emoji-rich, visual-first)
生成前加载:
  • references/platform-specs.md
    -- 轮播规格、Reel 尺寸、算法权重
  • references/hook-formulas.md
    -- 第一张幻灯片的 Hook、Caption 开头、Reel Hook
  • references/voice-adaptation.md
    -- Instagram 语气规则(轻松、Emoji 丰富、视觉优先)

Platform Rules

平台规则

SpecValue
Caption limit2,200 characters (125 visible before "more")
Carousel slides7-10 optimal, up to 20 max
Carousel dimensions1080x1350 (4:5)
Mixed media carousel2.33% engagement vs 1.80% image-only
Reel dimensions1080x1920 (9:16)
Reel sweet spot30-90 seconds
Algorithm priorityWatch time #1, then sends/shares, then saves
Hashtags3-5 niche (over-tagging reduces reach)
Best timeMon/Wed/Fri 11AM-1PM
规格值
Caption 上限2,200 个字符(点击“more”前可见 125 个字符)
轮播幻灯片最佳为 7-10 张,最多 20 张
轮播尺寸1080x1350(4:5)
混合媒体轮播互动率为 2.33%,纯图片轮播为 1.80%
Reel 尺寸1080x1920(9:16)
Reel 最佳时长30-90 秒
算法优先级观看时长第 1,其次是发送/分享,最后是收藏
Hashtags3-5 个细分话题(过度添加会降低触达率)
最佳发布时间周一/周三/周五 11:00-13:00

Output 1: Carousel Script (7-10 Slides)

输出 1:轮播脚本(7-10 张)

Write to
instagram/carousel.md
.
写入
instagram/carousel.md
。

Slide-by-Slide Structure

逐页结构

SlidePurposeContent
Slide 1Stop the scrollHook headline (5-8 words), visual pattern interrupt, "Swipe left" prompt
Slides 2-3Context + problemSet up why this matters, relatable pain point
Slides 4-6Core valueOne insight per slide, drawn from highest-impact atoms
Slides 7-8Proof + examplesStats, case studies, or before/after from atoms
Slide 9SummaryKey takeaway in one powerful sentence
Final slideTriple CTASave + Follow + Share prompts
页码目的内容
第 1 页阻止继续滑动Hook 标题(5-8 个词)、视觉模式中断、“向左滑动”提示
第 2-3 页背景 + 问题说明为什么这很重要,并引出 relatable pain point
第 4-6 页核心价值每页一个洞察,选取影响力最高的 atoms
第 7-8 页证据 + 示例来自 atoms 的统计数据、案例研究或前后对比
第 9 页总结用一句有力的话概括关键要点
最后一页三重 CTA收藏 + 关注 + 分享提示

Carousel Rules

轮播规则

  • Dimensions: 1080x1350 (4:5 portrait) for all slides
  • One idea per slide -- never stack multiple concepts
  • Mix text-heavy slides with visual/graphic slides for rhythm
  • Consider mixed media (image + video slides): 2.33% vs 1.80% engagement
  • Text must be readable at mobile size (large, bold headlines)
  • Consistent visual style across all slides (colors, fonts, layout)
  • Slide 1 determines whether anyone swipes -- invest the most effort here
  • 所有幻灯片尺寸均为 1080x1350(4:5 竖版)
  • 每页只讲一个观点——绝不堆叠多个概念
  • 交替使用文字较多的幻灯片与视觉/图形幻灯片,以保持节奏
  • 考虑使用混合媒体(图片 + 视频幻灯片):互动率为 2.33%,纯图片为 1.80%
  • 文字必须能在移动端尺寸下清晰阅读(使用大号、加粗标题)
  • 所有幻灯片保持一致的视觉风格(颜色、字体、布局)
  • 第 1 页决定用户是否继续滑动——应在这里投入最多精力

Visual Direction

视觉方向

For each slide, specify:
  • Headline text (5-10 words, bold)
  • Supporting text (1-2 short lines, if any)
  • Background treatment (color, gradient, image suggestion)
  • Layout (centered, left-aligned, split)
  • Any icon, illustration, or photo direction
为每页指定:
  • 标题文字(5-10 个词,加粗)
  • 辅助文字(如有,1-2 行短文)
  • 背景处理方式(颜色、渐变、图片建议)
  • 布局(居中、左对齐、分栏)
  • 任何图标、插画或照片方向

"Swipe Left" Prompt

“向左滑动”提示

Include a directional cue on slide 1. Only 5% of creators use this, but it boosts engagement by ~10%. Options:
  • Arrow graphic pointing right
  • Text: "Swipe for the full breakdown"
  • Visual element that extends past the right edge (implies continuation)
在第 1 页加入方向提示。只有 5% 的创作者使用这种方式,但它能提升约 10% 的互动率。可选方案:
  • 指向右方的箭头图形
  • 文字:“滑动查看完整拆解”
  • 延伸至右边缘之外的视觉元素(暗示内容尚未结束)

Output 2: Caption

输出 2:Caption

Write to
instagram/caption.md
.
写入
instagram/caption.md
。

Caption Structure

Caption 结构

  1. First 125 characters (visible before "more" fold)
    • This is the hook. Front-load the most compelling statement.
    • Use a curiosity gap, bold claim, or question
    • Must work as a standalone micro-copy even if nobody taps "more"
  2. Body (up to 2,200 chars total)
    • Expand on the carousel's theme with personal context or story
    • Use line breaks and emoji bullets for scannability
    • Weave in 1-2 atoms that complement (not duplicate) the carousel content
    • Conversational tone with contractions and natural language
  3. Engagement question
    • End with a specific question that invites comments
    • Questions that ask for personal experience outperform yes/no questions
  4. Hashtags (3-5)
    • Place at the very end after a line break
    • Mix 1-2 broad niche tags with 2-3 specific tags
    • Never use 30 hashtags (reduces reach in 2026)
  1. 前 125 个字符(在“more”折叠前可见)
    • 这是 Hook。将最有吸引力的陈述放在前面。
    • 使用好奇缺口、大胆声明或问题。
    • 即使没人点击“more”,也必须能作为独立的微文案发挥作用。
  2. 正文(总长度最多 2,200 个字符)
    • 围绕轮播主题展开,补充个人背景或故事。
    • 使用换行和 Emoji 项目符号,提高可扫描性。
    • 融入 1-2 个与轮播内容互补而非重复的 atoms。
    • 使用缩写、口语化且自然的语气。
  3. 互动问题
    • 以一个具体问题结束,邀请用户发表评论。
    • 询问个人经验的问题比是非题效果更好。
  4. Hashtags(3-5 个)
    • 另起一行,放在最末尾。
    • 混合使用 1-2 个较宽泛的细分标签和 2-3 个具体标签。
    • 不要使用 30 个 Hashtags(2026 年这会降低触达率)。

Caption Voice

Caption 语气

  • Relaxed and creative, not corporate
  • Emoji used as visual markers and personality (3-8 per caption)
  • Storytelling tone: conversational, flowing, first-person
  • 轻松且富有创意,不要像企业宣传
  • 将 Emoji 作为视觉标记和个性表达(每条 Caption 使用 3-8 个)
  • 叙事语气:对话式、流畅且使用第一人称

Output 3: Reel Script

输出 3:Reel 脚本

Write to
instagram/reel-script.md
.
写入
instagram/reel-script.md
。

Reel Structure (30-90 seconds)

Reel 结构(30-90 秒)

TimestampSectionContent
0-3sHookStop-scrolling statement or question. This determines everything.
3-10sContextQuick setup: "Here's what most people get wrong about [topic]..."
10-40sValueCore insight from 2-3 atoms. Break into 3-5 second segments with jump cuts.
40-55sProofStat, result, or example that validates the insight
55-70sCTATell viewers what to do: save, follow, comment, check link in bio
70-90sBonus (optional)Extra tip or teaser for next reel
时间戳部分内容
0-3sHook阻止滑动的声明或问题。这一点决定成败。
3-10s背景快速引入:“大多数人都会在[主题]上犯这样的错误……”
10-40s价值来自 2-3 个 atoms 的核心洞察。拆分为 3-5 秒的片段,并使用跳切。
40-55s证据用于验证洞察的统计数据、结果或示例
55-70sCTA告诉观众下一步操作:收藏、关注、评论或查看主页链接
70-90s额外内容(可选)额外提示或下一条 Reel 的预告

Reel Rules

Reel 规则

  • Dimensions: 1080x1920 (9:16 vertical)
  • Sweet spot: 30-90 seconds (shorter = better for discoverability)
  • Jump cuts every 3-5 seconds to maintain attention
  • Hook in first 3 seconds is non-negotiable (determines watch-through rate)
  • Watch time is the #1 algorithm signal -- every second must earn the next
  • 尺寸:1080x1920(9:16 竖版)
  • 最佳时长为 30-90 秒(越短越有利于内容发现)
  • 每 3-5 秒使用一次跳切,以保持注意力
  • 前 3 秒必须设置 Hook(决定完播率,这一点不可妥协)
  • 观看时长是算法的第 1 信号——每一秒都必须值得用户继续观看

Text Overlay Suggestions

文字叠加建议

For each section of the reel, suggest:
  • On-screen text (key phrase or keyword, 3-6 words)
  • Text position (top, center, bottom -- avoid bottom 20% for UI overlap)
  • Animation style (pop-in, typewriter, static)
为 Reel 的每个部分建议:
  • 屏幕文字(关键词或关键短语,3-6 个词)
  • 文字位置(顶部、居中或底部——避开底部 20%,以防与 UI 重叠)
  • 动画风格(弹出、打字机效果或静态)

Audio Direction

音频方向

  • Suggest whether to use: voiceover, trending audio, or original audio
  • If voiceover: the script IS the voiceover (write it conversationally)
  • Note: trending audio boosts discoverability but must match content mood
  • 建议选择:voiceover、热门音频或原创音频
  • 如果使用 voiceover:脚本就是 voiceover 内容(应使用对话式语言编写)
  • 说明热门音频虽然能提升内容发现率,但必须与内容情绪匹配

Image Prompts

图像提示词

If
--images
flag is set, suggest:
  • Carousel slides: 4:5 (1080x1350) visual concepts for each slide
  • Reel cover: 9:16 (1080x1920) thumbnail concept
  • Include /banana prompt descriptions for key slides
如果设置了
--images
标志,建议:
  • 轮播幻灯片:为每页提供 4:5(1080x1350)的视觉概念
  • Reel 封面:提供 9:16(1080x1920)的缩略图概念
  • 为关键页面包含
    /banana
    提示词描述

Output Format

输出格式

Each file should include the content plus production notes:
markdown
undefined
每个文件都应包含正文以及制作说明:
markdown
undefined

Instagram Carousel: [Topic]

Instagram Carousel: [Topic]

Slide 1 -- Hook

Slide 1 -- Hook

Headline: [text] Visual: [direction]
Headline: [text] Visual: [direction]

Slide 2 -- [Label]

Slide 2 -- [Label]

...

Posting notes: [best time, hashtag guidance, engagement strategy]
undefined
...

Posting notes: [best time, hashtag guidance, engagement strategy]
undefined

Brief Mode

简略模式

If
--brief
is active, produce only:
  • 1 caption (standalone, not tied to a carousel)
  • Skip carousel and reel script
如果
--brief
已启用,则仅生成:
  • 1 条 Caption(独立呈现,不依附于轮播内容)
  • 跳过轮播和 Reel 脚本