tuzi-copy-polish
Social Copy Polish
Optimizes social media copy for different platforms. Adjusts tone, length, hashtags, and formatting based on platform best practices.
Step 0: Load Preferences ⛔ BLOCKING
0.1 Check EXTEND.md
test -f .tuzi-skills/tuzi-copy-polish/EXTEND.md && echo "project"
test -f "$HOME/.tuzi-skills/tuzi-copy-polish/EXTEND.md" && echo "user"
| Result | Action |
|---|---|
| Found | Load, parse, apply settings |
| Not found | Continue with defaults (ask platform in Step 2) |
| Path | Location |
|---|---|
.tuzi-skills/tuzi-copy-polish/EXTEND.md |
Project directory |
$HOME/.tuzi-skills/tuzi-copy-polish/EXTEND.md |
User home |
EXTEND.md Supports: Default platform | Default language | Custom brand voice
Schema: references/config/preferences-schema.md
Step 1: Analyze Input
Determine input source:
- File path: Read file content as draft copy. Record the file path for Step 3 output.
- Inline text: Use the text directly as draft copy.
Identify:
- Content type (product promotion, knowledge sharing, personal story, news, opinion)
- Current language and tone
- Key message and selling points
- Target audience (if mentioned)
Step 2: Confirm Platform
If platform not specified by user or EXTEND.md, use AskUserQuestion:
header: "Target Platform"
question: "优化目标平台?"
options:
- label: "小红书"
description: "种草风、emoji、标签、口语化"
- label: "X/Twitter"
description: "简洁有力、英文友好、话题标签"
- label: "抖音"
description: "口语化、悬念开头、引导互动"
- label: "微信公众号"
description: "深度内容、段落清晰、引导关注"
- label: "通用"
description: "不针对特定平台,仅优化表达"
Step 3: Polish Copy
Load platform rules from references/platforms/, then rewrite:
Optimization Dimensions
| Dimension | Description |
|---|---|
| Hook | Opening line that grabs attention |
| Tone | Match platform culture (e.g., 小红书 casual, 公众号 professional) |
| Length | Trim or expand to platform sweet spot |
| Structure | Platform-specific formatting (line breaks, paragraphs, lists) |
| Hashtags | Add relevant tags per platform convention |
| CTA | Call-to-action matching platform interaction patterns |
| Emoji | Platform-appropriate emoji usage |
Output Format
If input is a file path: Append the polished copy to the end of the original file, separated by a clear divider:
---
## Polished Copy ({platform})
{polished content}
Then inform the user: "已将优化后的文案追加到文件末尾: {file_path}"
If input is inline text: Present the polished copy in a code block for easy copying.
In both cases, include a brief changelog (what was changed and why, 2-3 bullet points).
Multiple Variants
When the user asks for options, generate 2-3 variants with different angles:
- Variant A: Focus on emotion/story
- Variant B: Focus on value/benefit
- Variant C: Focus on curiosity/hook
Step 4: Iterate
After presenting the polished copy, ask if the user wants to:
- Adjust tone (more casual / more professional)
- Try a different angle
- Optimize for another platform
- Fine-tune specific parts
Extension Support
Custom configurations via EXTEND.md. See Step 0 for paths and supported options.
More from tuziapi/tuzi-skills
tuzi-short-video
为小红书、抖音、X/Twitter、视频号等平台生成短视频内容。分析用户输入,自动适配平台规格,生成视频脚本并调用视频生成后端。当用户要求"生成短视频"、"做一个视频"、"短视频"或为社交平台创建视频内容时使用。
161tuzi-image-gen
通过兔子API(nano-banana 模型)、Google、OpenAI、DashScope 和 Replicate 进行 AI 图片生成。支持文生图、参考图片、宽高比、模型选择,以及 OpenAI 兼容网关方言切换。当用户要求生成、创建或绘制图片时使用。
155tuzi-video-gen
通过兔子API进行AI视频生成。支持 Veo、Sora、Kling、Seedance 等模型,单视频和长视频(多段合成)模式。当用户要求生成视频、创建视频或需要视频生成后端时使用。
138tuzi-comic
知识漫画创作工具,支持多种画风和语调。创建原创教育漫画,包含详细分镜布局和顺序图片生成。当用户要求创建"知识漫画"、"教育漫画"、"传记漫画"、"教程漫画"时使用。
134tuzi-infographic
生成专业信息图,包含 21 种布局类型和 20 种视觉风格。分析内容,推荐布局×风格组合,生成可发布的信息图。当用户要求创建"信息图"、"可视化"、"高密度信息大图"、"visual summary"时使用。
134tuzi-post-to-wechat
通过 API 或 Chrome CDP 发布内容到微信公众号。支持文章发布(HTML、Markdown 或纯文本输入)和图文消息发布(多图)。当用户提到"发布公众号"、"微信公众号"、"贴图"、"图文"、"文章"时使用。
132