Skip to content
Back to search

handdraw-style-prompter

SkillWorks with Claude CodeGitHub

Turn a 001–370 hand-drawn and realistic style number and image theme into bilingual prompts, using model capability data to decide when core traits and a numbered reference image are required.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026. Its scripts were only partly scanned.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/yang0/handraw-style with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/yang0/handraw-style'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/yang0--handraw-style/badge.svg)](/listings/yang0--handraw-style)

HTML

<a href="/listings/yang0--handraw-style"><img src="/api/listings/yang0--handraw-style/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

直接把本仓库地址发给 Codex,让 Codex 自动完成安装:

> **“帮我安装这个 Skill:https://github.com/yang0/handraw-style”**

Codex 会自动拉取仓库并配置好全部风格与排版资源。

本项目聚焦手绘图片、封面、插图、IP、电商和摄影设计。视频制作、配音、分镜画廊和渲染已拆到独立的 `handraw-video` 项目,不再随图片 Skill 安装;视频封面设计作为静态图片能力保留。

其他功能的包内资源、运行工具与在线服务边界见[项目依赖说明](DEPENDENCIES.md)。提交前可运行`python scripts/audit_release.py`检查本机路径、常见凭据和视频文件是否混入。

本次自包含与隐私修正的范围、验证和历史提交边界见[发布检查记录](RELEASE_AUDIT.md)。

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

SKILL.md · Manifest · 236 lines
  1. ---
  2. name: handdraw-style-prompter
  3. description: Turn a 001–370 hand-drawn and realistic style number and image theme into bilingual prompts, using model capability data to decide when core traits and a numbered reference image are required.
  4. ---
  5. # Hand-drawn Style Prompter
  6. This is the install entrypoint for the complete hand-drawn style package.
  7. The package root contains the `images/` gallery and numbered reference assets;
  8. the operational contract is [skills/handdraw-style-prompter/SKILL.md](skills/handdraw-style-prompter/SKILL.md).
  9. Before handling a request, read that Skill file completely. Resolve every
  10. gallery and reference-image path from this installed package root. Install this
  11. repository at path `.` so the `images/` directory and the nested Skill are
  12. kept together.
  13. Read [package dependencies](DEPENDENCIES.md) before claiming that an installed
  14. Skill can generate or render media. Bundled source/resources and separately
  15. installed tools/services are distinct; never rely on the author's workstation.
  16. Video production, narration and rendering are maintained in the separate `handraw-video` project. This package remains focused on static images, including article and video cover images.
  17. ## Post-generation sharing across all routes
  18. Before parsing any new creative request, capture the user's original creative
  19. message verbatim, including whitespace and line breaks, in `original-request.txt`
  20. inside the local output directory. Follow-up selections and enlargement commands
  21. inherit the original request for that creative task. Do not substitute generated
  22. prompts, conversation history, environment context or credential messages.
  23. After selecting the actual style and layout, record their canonical IDs. Public
  24. sharing keeps the original request verbatim and appends only compact actual-ID
  25. tags, such as `【画风:FA-013 】` and, only when used, `【图型:IG-038 】`.
  26. Omit note headings, explanations and unused/unrecorded fields. Supply IDs with
  27. `--style` / `--layout` (repeat for multiple IDs), or `--no-style` / `--no-layout`
  28. when unused. Missing historical IDs mean unrecorded, not unused. Do not infer
  29. IDs from generated images or include intermediate/expanded prompts.
  30. The share destination is the website's work square: create a free `kind=work`
  31. entry through `/studio/api/entries`, with actual style/layout asset references.
  32. Do not substitute a forum post or return a forum topic URL as a work receipt.
  33. After delivering verified generated images from any routed Skill, follow
  34. [conversation sharing](skills/community-share/references/conversation-sharing.md)
  35. to prepare a local draft and show the returned **分享到网站** Markdown link.
  36. The link opens a loopback preview page with a real HTML publish button in local
  37. Codex and Claude clients. Do not assume host directives render as buttons. When
  38. a local service cannot run, offer an honest text trigger instead. Preparing a
  39. draft and opening its preview never publish. Publish only after the user clicks
  40. the page's publish button or explicitly requests sharing.
  41. Resolve these resources from this installed package root, not a flattened
  42. `.agents/skills` entry. Put the share action before the required gallery footer.
  43. ## Request routing index
  44. When the user asks to illustrate an article, plan article illustrations, choose
  45. insertion points, or write prompts for article images, read and invoke
  46. [article-illustration-planner](skills/article-illustration-planner/SKILL.md)
  47. before handling the request. Use the hand-drawn style rules in
  48. [handdraw-style-prompter](skills/handdraw-style-prompter/SKILL.md) for style-number
  49. resolution and prompt construction.
  50. When the user asks to design a poster, generate poster prompts, or create structured poster prompts, read and invoke
  51. [poster-prompt-generator](skills/poster-prompt-generator/SKILL.md), actively recommending cohesive style and theme color pairings from the visual style library (#001–#370) and theme color library (C-01–C-36).
  52. When the user asks to design an article cover, video cover, generate cover prompts, create a WeChat official account cover, design an X/social cover for an article, or design a video cover from subtitle files/transcripts (例如“给文章做封面”、“文章封面设计”、“视频封面设计”、“视频封面”、“字幕做封面”、“公众号封面”、“X文章头图”), read and invoke
  53. [article-cover-designer](skills/article-cover-designer/SKILL.md) before handling the request. It extracts a ~200-word summary, infers target audience, sets the business scenario (article or video cover), injects style (#001–#370) and theme color (C-01–C-36), and applies the fixed prompt instruction ("先设计隐喻再出图,主标题明显,小字少或者没有小字。 其他你帮我设计。";若是视频封面或输入字幕文件,自动注入“视频封面设计,网感强,吸引人眼球”)。
  54. When the user asks to blend, merge, or fuse two styles, generate a cross-media fusion prompt, or create a "character visual language × scene visual language" artwork (例如“【风格融合设计】”、“风格融合设计”、“风格融合”、“融合两种风格”、“角色与场景风格融合”、“双风格融合”、“跨媒介风格融合”), read and invoke
  55. [style-fusion-prompter](skills/style-fusion-prompter/SKILL.md) before handling the request. It supports fusing two style numbers (#001–#370), or a style number with a realistic style, optionally injects layouts (SC-*/IG-*) and theme colors (C-01–C-36), and applies the strict dual visual language coexistence template.
  56. When the user asks to design an IP character, build a Character Bible or IP Design Manual, generate standard IP cards (例如“IP形象设计”、“角色设计手册”、“设计三视图与周边”、“IP-01至IP-18卡片”), read and invoke
  57. [ip-designer](skills/ip-designer/SKILL.md) before handling the request. It supports 5 layers (Strategy, Character, Visual, System, Commercial) across 18 standard cards (6-card basic, 10-card standard, or 18-card full edition).
  58. When the user asks to manage, add, query, replace, or delete assets in the custom library (例如“添加到角色库”、“保存到道具库/场景库”、“替换角色基准图”、“修改资产名称/标签”、“初始化/挂载图库目录”、“查看图库清单”), read and invoke
  59. [custom-asset-manager](skills/custom-asset-manager/SKILL.md) before handling the request. It manages the full lifecycle of custom character (CH-*), prop (PR-*), and scene (SCN-*) assets with zero Git conflicts.
  60. When the user asks to plan a photo shoot, create a photography project, generate shot lists, design couple/wedding/family/portrait/cosplay photography plans, or draw card for a themed shoot (例如“拍摄企划”、“情侣照”、“婚纱照”、“写真”、“全家福”、“Cosplay拍摄”、“抽卡”、“8宫格”、“摄影分镜”), read and invoke
  61. [couple-photo-orchestrator](skills/couple-photo-orchestrator/SKILL.md) before handling the request. It orchestrates the full pipeline from direction selection through Look design, scene micro-zones, pose-emotion planning, Share Card compilation, and 8-shot adaptive grid generation.
  62. When the user asks to share creations, post to any forum category, publish images or bilingual prompts to the community/forum, login/bind website account, view published posts, edit/delete community topics, deploy or sync static gallery pages and assets to the cloud, or update the website (例如“发到论坛”、“社区发帖”、“手绘风发帖”、“发到创作交流/好用的工具/手绘skill/综合大厅”、“分享作品到社区”、“查看我的帖子”、“删除我的帖子”、“绑定手绘风账号”、“发布静态页面到社区”、“同步画廊”、“更新网站”、“部署静态页”), read and invoke
  63. [community-share](skills/community-share/SKILL.md) before handling the request. It provides a dual publishing engine for both Discourse forum interaction (ID 4–9) and static gallery cloud deployment (syncing HTML, 365 styles, and WebP assets to tencent-vm / CDN with optional forum update announcements).
  64. When the user asks to add or import a named or text-defined visual style to the style library, register positive visual traits, or generate representative sample tiles for a new style (例如“添加画风”、“录入风格”、“新增手绘风格”、“自定义风格入库”), read and invoke
  65. [style-library-importer](skills/style-library-importer/SKILL.md) before handling the request. It processes style names and filtered positive traits, generates clean representative images, assigns the next style number, and updates metadata and contact sheets. 严格遵守特征提取铁律:严禁从示意图/参考图中解析提取任何具体人物、服装、动作、道具或场景地貌写入特征(traits),特征只保留纯粹的媒介、光影、色彩与艺术视觉语言。
  66. When the user supplies an X/Twitter tweet link and asks to scrape, import, or convert tweet AI styles or `--sref` images into the numbered style library (例如“抓取推文画风”、“推文风格入库”、“导入Twitter风格”、“把这条推特录入到风格库”), read and invoke
  67. [tweet-style-importer](skills/tweet-style-importer/SKILL.md) before handling the request. It automatically extracts full-resolution images, generates 4-grid reference and clean single tiles, configures mandatory reference passing, and rebuilds the library. 严格遵循内容解耦规范,严禁将推文原图的画面具体主体内容写入特征描述。
  68. When the user asks to add a new visual layout or composition structure to the layout gallery (such as social cards, infographics, comic storyboards, IP cards, or e-commerce layouts; 例如“添加排版图型”、“录入新版式”、“导入排版”、“新增Layout”), read and invoke
  69. [layout-library-importer](skills/layout-library-importer/SKILL.md) before handling the request. It losslessly converts the reference image to WebP, records category metadata, creates bilingual prompt files, and updates the layout gallery.
  70. When the user asks to analyze the commercial application of a style, write or update style business guides, plan business use cases/products, or audit missing business documents (例如“商业应用”、“风格商业分析”、“商业指南”、“适合什么生意”、“商业落地”、“变现策划”、“business指南”、“补充商业文档”), read and invoke
  71. [style-business-analyzer](skills/style-business-analyzer/SKILL.md) before handling the request. It authors gold-standard 5-module business guides in `business/{number}.md` with deep insights, product portfolios, and layout/color synergy.
  72. ## Default mode
  73. Default to creating prompts only. Do not call an image-generation tool unless the user explicitly asks to generate, render, or preview an image.
  74. ### Automatic Style and Theme Color Recommendation Policy (默认智能推荐风格与主题色)
  75. 不管生成什么图片或提示词,**默认情况下如果用户没有指定风格和主题色,必须自动推荐最优的风格与主题色组合**,严禁因用户缺少编号而中断或拒绝生成:
  76. > [!IMPORTANT]
  77. > **【核心准则:每次由 AI 实时动态判断,严禁写死/机械套用】**
  78. > - **实时动态研判**:严禁在记忆、规则或提示词中维护任何“主题关键词 -> 固定风格编号/颜色编号”的静态死板映射(例如严禁一见节气/传统就推 268、一见自然就推 266、一见科技就推 011/054、一见生活就推 018)。
  79. > - **全库开放式匹配**:每次推荐必须由 AI 结合当前用户具体主题的深层语境、精神内核、视觉隐喻与画面构图,在全库 365 种视觉风格(#001–#370)与 36 种经典主题色(C-01–C-36)中进行**实时、开放式的审美推理与动态搭配**。
  80. > - **激发全库多样性**:同一个主题在不同设计视角下具备多维的美学可能性(如“菜园”既可以是水墨写意、也可以是田园木刻版画、粗粒油画棒、或包豪斯几何构成)。每次推荐都应根据具体切入点构思,充分展现全库 370 种画风与 36 种色彩的丰富生命力。
  81. > - **美学理由具象化**:AI 给出的美学推荐理由必须紧扣当前主题的视觉隐喻和画面构图,说明为什么该画风的笔触/质感与该色彩的情绪能完美传达这一主题,杜绝套话。
  82. 1. **风格与主题色均未指定**:根据用户输入的主题语义、情感基调、受众与使用场景,由 AI 实时动态从 365 种视觉风格与 36 种经典单色库中推荐 1 组契合度最高的【风格编号 (#001–#370) + 主题色编号 (C-01–C-36)】组合。主题色数量不设死限,依据画面层次灵活决定单色统领、双色搭配(主色+点缀色)或三色调和,简述 1 句具象化美学推荐理由,并直接输出完整生图提示词(或执行出图)。
  83. 2. **仅指定风格,未指定主题色**:严格保留用户指定的风格,根据该风格与画面主题,由 AI 实时动态推荐最协调的【主题色 (C-01–C-36)】搭配(可为单色或多色组合)。
  84. 3. **仅指定主题色,未指定风格**:严格保留用户指定的主题色,根据色彩调性与画面主题,由 AI 实时动态推荐 1 款最契合的【视觉风格 (#001–#370)】搭配。
  85. 4. **两者皆指定**:完全遵照用户指定的内容输出。
  86. ## Style activation policy
  87. Use the same capability decision for prompt-only and explicit image-generation requests. Resolve the selected/current model (or model family) against `skills/handdraw-style-prompter/references/model_capabilities.json`; if no model is specified, default to `gpt-image-2` (the calibrated baseline model of this library). Only use the default unknown capability fallback when the user explicitly requests an uncalibrated third-party model. The decision uses the indexed author name plus the generated style name and is not based on the author's fame or life status.
  88. - `name_activation=strong`: use only the indexed author name, generated style name, and theme. (Specifically for styles like #042 Beatrix Potter, #011 David Shrigley, #269 Socialist Realist Propaganda Poster, and other strong-activation styles: the model understands the aesthetic natively; output only the reference author and style name, and STRICTLY DO NOT output core traits, local image file paths, or reference-isolation blocks.)
  89. - Otherwise, include every available positive core trait with the author/style name and theme.
  90. - If name plus traits is not strongly activated, also require the configured reference asset. For explicit generation, pass it through `referenced_image_paths`. In `pure-image` prompt-only output, write the local asset path and reference-isolation instruction inside both prompts for the user to upload manually. In `graphic-text` prompt-only output, do not place a path, upload instruction, or isolation block inside either copyable prompt; show the resolved reference image to the user outside the prompts instead. Assets live under the installed package root in numbered 200-style buckets: for example, #217 uses `images/individual/201-400/217_grid.webp`.
  91. - If no positive core trait exists and name is not strongly activated, use the author/style name, theme, and reference image.
  92. - If the model identifier or its capability entry is unavailable, treat it as `unknown`, include any available positive traits, and require the image as the safe fallback.
  93. - Use `python skills/handdraw-style-prompter/scripts/resolve_reference.py --model <model> --style <number>` when a deterministic decision check is useful. The script prints JSON and never guesses an unknown model's capability.
  94. - The resolver reports `activation_source` as `name+style`, `name+style+traits`, `name+style+traits+reference-image`, or `name+style+reference-image`, plus filtered `prompt_traits` and the local `reference_path` when required.
  95. - Resolve the style number to its configured reference asset relative to the installed package root. The normal fallback is `images/individual/{bucket}/{number}.webp` (for example, `048` maps to `images/individual/001-200/048.webp`); a matching `{number}_grid.webp` in the same bucket takes priority.
  96. - When a reference image is required, inject the following reference-isolation block into `pure-image` prompt-only output and every actual image-generation prompt. Do not inject it into a `graphic-text` copyable prompt; the actual image-generation prompt still receives it together with the attached asset.
  97. Chinese: `所附图片仅用于参考画风。只提取参考图的风格特征,例如线条、笔触、媒介、材质、色彩倾向和整体视觉语言;不要使用、复制或延续参考图中的任何主体、人物、动物、服装、道具、动作、姿态、场景、背景、构图、布局、文字或故事。最终画面内容完全以用户提供的主题为准。`
  98. English: `Use the attached image only as a style reference. Extract only its stylistic qualities, such as linework, brushwork, medium, material texture, color tendencies, and overall visual language. Do not use, copy, or carry over any subject, person, animal, clothing, prop, action, pose, setting, background, composition, layout, text, or story from the reference image. The user's written theme is the sole source for the image content.`
  99. - The user's theme is the sole source for subjects and narrative; the reference image must never override or add content to the theme.
  100. - Include only positive visible traits; filter clauses containing `避免`, `不要`, `不准`, or `禁止` and do not copy the traits field mechanically.
  101. - If the numbered image is missing or cannot be passed, report that limitation and provide the normal text prompts; never invent or substitute a reference image.
  102. - After generation, identify whether the numbered reference image was used. If used, identify its number. Do not imply generation when the user requested prompts only.
  103. ## Session initialization
  104. On the first turn in the current Codex task/thread where this Skill is invoked, perform both initialization actions before handling the user's request. This applies to **任何首次请求**, including opening the gallery, browsing the index, requesting prompts, or requesting image generation:
  105. 1. If this task has not already displayed the gallery, dynamically resolve the absolute `file://` URI to `skills/handdraw-style-prompter/gallery/index.html` based on the current workspace or installed package root, and call `mcp__codex_app__open_in_codex` in a browser tab with `target.type="browser"` and that URL. Never open `gallery/index.html` as `target.type="file"`, in an editor, or as a file preview.
  106. 2. Include the exact standalone status line `当前处于纯图模式,可切换为图文模式。` in the same final reply, even when the request produces no prompt.
  107. 3. Continue with the user's request after the browser call; opening the gallery must not block prompt generation or an explicitly requested image-generation follow-up. If the same first reply would also add the pure-image mode note, show this status line only once.
  108. 4. Do not repeat the browser call or this first-session status notice on later turns in the same task/thread. Use the conversation context (not a persistent state file) to determine whether initialization already happened.
  109. 5. If the browser call is unavailable or fails, provide the dynamically resolved local `file://` fallback link, then continue normally.
  110. This initialization applies only when this Skill is invoked for the first time in a task/thread; unrelated conversations must not open the gallery.
  111. ## Inputs
  112. For style-only work, if the style number (`001`–`370`) is omitted, automatically recommend an optimal style and theme color combination based on theme semantics. For layout work, require a layout ID (`SC-001` or `IG-001`); if style and/or theme color are omitted, automatically recommend harmonious ones. Accept optional theme color (`C-01`–`C-36` or color name), aspect ratio, subject constraints, text requirements, and a mode. When the user requests a prompt (e.g. specifying a style and/or layout) without providing a theme, do NOT pause or halt to ask for a theme—immediately output the complete prompt with a clear placeholder (such as `【请在此输入画面主题,或在生图模型中垫入你的照片】` / `[Enter theme here, or attach your photo in the image AI]`). If a supplied number, layout ID, or color ID is invalid, ask the user to choose a valid indexed value; do not invent one. Do not add an aspect ratio when none was supplied.
  113. Users can browse `skills/handdraw-style-prompter/gallery/index.html` for numbered style contact sheets, `skills/handdraw-style-prompter/gallery/layouts.html` for layout thumbnails, and `skills/handdraw-style-prompter/gallery/colors.html` for classic monochrome theme colors. The authoritative style content is `styles_200_reorganized.md`; `skills/handdraw-style-prompter/references/styles.json` is a generated index and must be refreshed with `python skills/handdraw-style-prompter/scripts/build_library.py` after Markdown changes. Layout metadata is `skills/handdraw-style-prompter/references/layouts.json`; each entry's bilingual prompt file under `skills/handdraw-style-prompter/references/layouts/` is the authoritative layout content and the layout gallery is refreshed with `python skills/handdraw-style-prompter/scripts/build_layout_gallery.py`. Monochrome color metadata is `skills/handdraw-style-prompter/references/colors.json` and refreshed with `python skills/handdraw-style-prompter/scripts/build_color_gallery.py`.
  114. ## Layout workflows
  115. - A layout ID selects composition and text structure, not illustration style. Use its prompt file verbatim in the selected output language, then append the user's theme as the only source of subject matter and copy.
  116. - **排版图型的两大架构分类 (Dual Architecture of Layout Patterns)**:
  117. 本库全部 170 种排版图型在底层设计与执行机制上明确划分为两大类:
  118. 1. **确定性静态排版(纯文本直接拼接型 · Static Templates)**:
  119. - 包括绝大多数常规图型(如 `SC-001`~`SC-023` 社媒卡、`IG-001`~`IG-035` 信息图、`SB-001`~`SB-071` 漫画分镜、`IP-001`~`IP-013` IP设计、`EC-001`~`EC-024` 电商)。
  120. - 构图拓扑关系单一固定,生图模型能直白无误地理解。处理方式为拿来即用,直接提取其 prompt 模板并拼接指定画风与主题输出。
  121. 2. **高维动态解析型排版(Skill 级动态解析决策型 · Generative Frameworks)**:
  122. - 代表图型为 `SC-021`(自适应双拼照片转译社媒卡)。
  123. - 该类图型不是单一路线的固定线框,而是高度抽象的“跨媒介视觉构图与空间语法系统”。生图模型不具备自主解析多选菜单的能力,若直接平铺备选清单,模型会因语义冲突退化为机械照搬原片全景背景。
  124. - **执行机制**:必须由 AI 助手充当设计总监(Art Director)进行前置解析与动态决策:
  125. - **比例驱动切分原则(严禁死板切分,完全由目标画幅比例决定)**:竖版画幅(如 3:4、9:16、1:2 等)默认采用上下 1:1 纵向堆叠双拼(上方摄影/下方手绘),顺应手机端自上而下的视觉动势与故事自然向下延展;横版画幅(如 16:9、4:3、5:2、21:9 等)默认采用左右 1:1 横向并排双拼(左侧摄影/右侧手绘),呈现独立艺术杂志跨页质感;方版画幅(1:1)自适应切分或服从用户显式指定;
  126. - **七大核心转译类型体系(Type 01 ~ Type 07)**:若用户未显式指定构图类型,由 AI 结合画面主题推荐最契合的 1 种类型(或按顺序轮替抽卡),并为选定的唯一类型生成单义、排他的执行提示词:
  127. * `【类型 01 · 视线锚点跨界续绘型 (Visual Anchor Continuation)】`:单一核心动态锚点(水/路/光/手势)跨越中缝边界流淌进入暖象牙纸手绘区域,在下方衍生绘本互动故事与极简黑线小人;
  128. * `【类型 02 · 撕纸裂变第二世界型 (Torn-Paper Second World)】`:不规则手工物理撕纸毛边接缝,现实穿透撕纸后物理规则超现实突变(水变布匹/光可手捧/影子成阶梯),暖象牙纸与克制拼贴;
  129. * `【类型 03 · 越界破框型 (Boundary Break / Out-of-Bounds)】`:几何色块/容器承托,主体关键部位(头/手/枝叶)突破色块边缘伸出,形成空间穿透感;
  130. * `【类型 04 · 微缩浮岛型 (Miniature Diorama / Floating Island)】`:主体微缩为 10%–25% 纸面小景,单抹色带或微基座承托,四周大片静谧空旷留白;
  131. * `【类型 05 · 记忆图谱平铺型 (Memory Knolling / Fragment Collage)】`:居中放置手绘主体,留白中错落平铺 3–6 枚带 NO. 编号的原图记忆特征贴纸;
  132. * `【类型 06 · 经典悬浮落地型 (Suspended / Grounded Hero)】`:主体尺度适中(占 40%–55%),底部仅用一条极简手绘地面摩擦线提示空间支撑;
  133. * `【类型 07 · 正负形咬合型 (Gestalt Positive-Negative Interlock)】`:主体高度提炼为几何剪影,与背景正负形相互转换咬合;
  134. - **背景强力净化指令**:坚决剔除原片石墙、马路、树木等环境杂质,确保艺术转译侧四周 40%–80% 纯净负空间与呼吸边距;
  135. - **强制交互点选输出规范**:每当使用或输出 SC-021 提示词时,在提示词文本末尾,**必须固定附带 7 大可选类型索引清单**(标明当前已采用类型,并提示用户回复编号如“换成 02”即可一键换装),赋予用户随时点选调整的机会。
  136. - **传图与改图规则(无需强制传图,完全支持用户传图与改图)**:无论使用 `SC-021`(自适应双拼照片转译社媒卡)还是其他任何风格/图型,用户传图均属**可选行为**:
  137. - **用户未传图时**:完全支持纯文本直出。AI 助手直接将排版规则与指定风格特征拼接为完整提示词(若未提供主题则提供直观占位符),绝不阻断对话或强制要求用户必须传图,由生图模型原生自主生成写实画面与风格转译。
  138. - **用户传图 / 改图时(写实改手绘 / 垫图转风格)**:完全支持!当用户上传照片或参考图要求“改图”、“改风格”、“转手绘”时,为了防止生图模型在写实照片表面机械叠加滤镜,提示词中**必须使用以下固定转绘重绘指令**:
  139. `【请只提取人物的五官特征和姿态,场景轮廓,在画风上严格按 选择的风格重新画,不要在原照片上加质感。】`
  140. (英文对应:`[Extract only the character's facial features and posture, and the scene contours. Strictly redraw from scratch according to the selected style; do not add texture or filters onto the original photo.]`)
  141. AI 助手提取参考图的核心特征(人物外貌特征、姿态、服装轮廓、场景概貌)作为主题描述,并无条件附带上述固定指令,强制生图模型在手绘维度从零重新绘制,彻底杜绝写实滤镜与照片微表面保留。
  142. - 核心原则在于:最终的像素级视觉绘制与画风转译始终由生图模型(如 GPT Image、Midjourney、Flux 等)原生完成。
  143. - When a layout ID is selected, automatically use graphic-text mode even if the user did not name a mode (except for E-commerce layouts `EC-*`: e-commerce layouts do not append the generic graphic-text suffix, and do not stitch a visual style unless explicitly requested by the user). For standard layouts, the layout prompt and the fixed graphic-text suffix are additive (stacked): append the fixed generic graphic-text suffix verbatim at the end of the prompt to combine concrete layout rules with graphic-text semantic guidance. In layout requests, never append the pure-image mode notice `当前处于纯图模式,可切换为图文模式。`, and never include reference-image file paths or upload instructions inside the copyable prompt.
  144. - If the user also supplies a style number, combine the selected style's author/style label, permitted positive traits, and reference-image policy with the layout prompt (for e-commerce layouts, only stitch style if the user explicitly specified a style number). A style reference image may influence only rendering style; it must not override the selected layout, text structure, or theme.
  145. - Return a single complete layout-combination prompt in the language of the user's theme. Chinese characters select Chinese; otherwise use English. Treat mixed input as Chinese. Existing style-only requests retain their normal bilingual output.
  146. - The layout gallery is browse-only: it has no prompt input form. It shows a numbered thumbnail, opens a large preview on click, and copies the canonical Chinese layout prompt on request.
  147. ## Prompt modes
  148. Default to `pure-image`. Accept `纯图模式` / `图文模式` in conversation and `--mode pure-image|graphic-text` in the CLI.
  149. - `pure-image`: preserve the current prompt workflow. After the normal reply, add the small note `当前处于纯图模式,可切换为图文模式。`.
  150. - `graphic-text`: preserve the user's theme exactly after `主题:` / `Theme:`. Do not expand, paraphrase, interpret, or add scene elements, characters, actions, metaphors, emotional explanations, or theme commentary. Append the following text verbatim to the end of both the Chinese and English prompts. Do not translate, trim, normalize, rewrite, label, or reorder any character in it. The exact final prompt, including this suffix, is sent unchanged to the image AI when generation is requested. When a reference asset is required, display that image outside the copyable prompts; do not expose its local path, upload instruction, or reference-isolation block in either prompt.
  151. `【如果主题直白包含画面元素那就按主题出图,文案由你来升华,但是不要直接描述画面。 如果主题比较概念化,那么文案和主题尽量保持一致,如果文案较长由你提炼,由你先设计画面隐喻(人类和非人类都行)再出图 。 文字参与构图,图文一体】`
  152. ## Output
  153. For a valid request, return these parts:
  154. 0. (若触发智能推荐) 推荐组合说明:
  155. - 标注推荐的风格与主题色(例如:`💡 智能推荐组合:风格 #{编号} · {generation_name} + 主题色 {id} · {name_zh} ({name_en})`);
  156. - 给出 1 句简明美学推荐理由;
  157. - 提示用户若有偏好的其他编号可随时说明替换。
  158. 1. Selected style: number and generated style name. Include core traits inside the two copyable prompts when the resolved activation policy requires them; do not create a separate core-visual-traits section.
  159. 2. Chinese prompt: begin the copyable prompt itself with `风格名称:{generation_name}。`(重要规则:生成的提示词内容内部严禁包含任何编号,包括风格编号如 `#018`、图型编号如 `SC-021` / `SC-001`、色彩编号如 `C-01` 等,直接使用自然语言名称,避免生图模型误将编号代码作为文字绘制到画面中);when a theme color is applied, include `主题色:{name_zh}({name_en})。`(不含色彩编号);describe only the user's theme and constraints explicitly provided by the user, and always include the indexed author/style name as a short `参考作者/风格名称` label. When required, append filtered positive traits as `核心风格特征:...`;当用户要求改图/垫图转画风时,在画面要求后必须追加:`【请只提取人物的五官特征和姿态,场景轮廓,在画风上严格按 选择的风格重新画,不要在原照片上加质感。】`(其中“选择的风格”可根据所选风格名称具体实例化)。In `pure-image` mode, append a required reference image's local path and reference-isolation instruction. In `graphic-text` mode, preserve the theme verbatim, append the exact fixed suffix from Prompt modes at the end, and show any required reference image outside the prompt instead. Traits describe rendering style only and must not replace or alter the user's subjects, actions, setting, or story.
  160. 3. English prompt: begin the copyable prompt itself with `Style name: {generation_name}.` (Strictly omit any numeric or alphanumeric IDs such as `#018`, `SC-021`, or `C-01` from the prompt text); when a theme color is applied, include `Theme color: {name_en}.`; describe only the same theme and user-provided constraints, and always include the indexed author/style name as `Reference author/style name`. When required, append the same traits as `Core style traits: ...`; when the user requests an image redraw/style transfer from an uploaded photo, append: `[Extract only the character's facial features and posture, and the scene contours. Strictly redraw from scratch according to the selected style; do not add texture or filters onto the original photo.]` (where "selected style" may be replaced by the chosen style's English name). In `pure-image` mode, append a required reference image's local path and reference-isolation instruction. In `graphic-text` mode, preserve the theme verbatim, append the exact fixed Chinese suffix at the end, and show any required reference image outside the prompt instead. Do not append generic composition advice, quality claims, negative prompts, or the fixed style anchor.
  161. 4. A brief note: the prompt can be pasted into any image AI; generation is controlled by that AI.
  162. Do not invent visual traits, extra style descriptions, generic quality/composition language, or default avoid-list wording. Include each entry's original reference author/style name from the index in both prompts as requested; this is an index label, not a claim about the person or an instruction to imitate them. Never use fame or life status as a proxy for model capability.
  163. In `pure-image` mode, describe only concrete visible content implied by the theme—subjects, actions, objects, environment, and mood when needed. In `graphic-text` mode, use the user's theme verbatim and leave all semantic expansion to the fixed suffix. In both modes, leave composition, layout, visual richness, quality, and rendering decisions to the image AI unless a layout ID was explicitly selected. Respect a user-specified text requirement but do not invent copy.
  164. ## Utilities
  165. - From the installed package root, rebuild the derived index and gallery: `python skills/handdraw-style-prompter/scripts/build_library.py`
  166. - Rebuild the layout gallery: `python skills/handdraw-style-prompter/scripts/build_layout_gallery.py`
  167. - Rebuild the color gallery: `python skills/handdraw-style-prompter/scripts/build_color_gallery.py`
  168. - Split contact sheets into numbered single images: `python skills/handdraw-style-prompter/scripts/split_contact_sheets.py`
  169. - Validate all source/index/gallery invariants: `python skills/handdraw-style-prompter/scripts/validate_library.py`
  170. - Produce a deterministic CLI prompt draft: `python skills/handdraw-style-prompter/scripts/prompt_style.py --style 18 --theme "秋天的第一杯奶茶"`
  171. - Produce a layout-combination draft: `python skills/handdraw-style-prompter/scripts/prompt_style.py --layout SC-001 --style 18 --theme "秋天的第一杯奶茶"`
  172. - Produce a color-combination draft: `python skills/handdraw-style-prompter/scripts/prompt_style.py --style 18 --color C-01 --theme "秋天的第一杯奶茶"`
  173. - Resolve image-reference policy: `python skills/handdraw-style-prompter/scripts/resolve_reference.py --model <model> --style 18`
  174. The CLI is a convenience check. For normal conversational use, write natural bilingual prompts rather than echoing its template mechanically.
  175. ## Personal resources and community collections
  176. When the user asks for 我的画风、我的图型、我的收藏、同步收藏、个人画风/图型 or importing a community resource package, read [personal-library](skills/personal-library/SKILL.md). Official entries remain unchanged; user-created and community-acquired entries live in the separate user library. Never upload personal resources without an explicit sharing request.
README.md · README · 292 lines
  1. <p align="center">
  2. <strong>中文</strong> | <a href="README_en.md">English</a>
  3. </p>
  4. # 手绘风格与排版图型提示词库(Hand-drawn Style & Layout Prompter)
  5. <p align="center">
  6. <a href="https://hellogithub.com/repository/yang0/handraw-style" target="_blank">
  7. <img src="https://api.hellogithub.com/v1/widgets/recommend.svg?rid=da31f932a6ef446fa2608d20f09c9be9&claim_uid=8adohGgCW07fknL" alt="Featured|HelloGitHub" style="width: 250px; height: 54px;" width="250" height="54" />
  8. </a>
  9. </p>
  10. > **不会描述画风?排不好版面构图?选个编号,一键生成极具辨识度的 AI 生图提示词。**
  11. 本项目整理了 **001–365 种手绘与写实视觉风格**、**170 种排版图型**(社媒卡、信息图、漫画分镜、IP设计、电商)与 **36 种经典单色主题色**。
  12. 不管是做小红书爆款图文、公众号题图、知识科普长图,还是绘制多格叙事条漫,你都不需要背诵复杂的画风术语,也不需要费心构思版面结构——**只要选定风格编号、排版编号与主题色,告诉 AI 你想表现的主题,就能立即获得经过实测验证的中英双语生图提示词**,直接复制到支持生图的 AI 工具中出图。
  13. > [!NOTE]
  14. > ### 📢 开源许可与致谢说明
  15. > **欢迎自由使用、修改和商业化。**<br>
  16. > 唯一的小请求:如果这个项目对你有帮助,或者你的项目基于它进行了二次开发,**请保留作者 yang0 和原仓库地址**([https://github.com/yang0/handraw-style](https://github.com/yang0/handraw-style))。开源不易,感谢你的理解与支持!
  17. > [!TIP]
  18. > ### 📚 官方飞书知识库已上线
  19. > 🔗 **[👉 点击进入官方飞书知识库 ↗](https://ecmn3m3i17.feishu.cn/wiki/space/7689222460876524523?ccm_open_type=lark_wiki_spaceLink&open_tab_from=wiki_home)**
  20. >
  21. > 1. 🖼️ **速查画廊**:移动端与网页端秒级检索 365 种视觉风格与 170 种排版效果大图
  22. > 2. ❓ **Skill 常见问题与提示词速查**:安装配置、出图答疑、常见生图报错排查与经典提示词模版
  23. > 3. 💡 **创业与变现经验**:持续更新社群创作者分享的一线实操变现案例与商业探索心得
  24. ---
  25. ## 🌟 核心解决的问题与特色
  26. | 创作痛点 | 本库的解决方案 |
  27. | :--- | :--- |
  28. | **画风描述模糊,每次生成都漂移** | **编号化精准激活**:365 种手绘插画与写实摄影风格体系(支持 FA~FI 分类编号与 001–365 传统编号),告别“可爱一点、文艺一点”的玄学生图。 |
  29. | **版式构图单调,排不好社媒长图** | **170 种排版图型**:24 种社媒卡、38 种数据信息图、71 种漫画分镜、13 种IP设计、24 种电商图型,构图版式开箱即用。 |
  30. | **色彩杂乱,缺乏统领画面的基调** | **36 种经典单色库**:克莱因蓝、鼠尾草绿、爱马仕橙、佩恩灰等,一键赋予画面高级纯正色调。 |
  31. | **图文不搭,文字塞不进画面** | **双模式随心切换**:支持“纯图模式”(纯插画表现)与“图文模式”(文字与画面一体化构图)。 |
  32. | **模型画风锁死或无法识别** | **模型能力分层与兜底**:显式模型精准标定(关键词/特征激发),通用与第三方模型自动参考图兜底保真。 |
  33. | **文章封面难出彩,文案画面脱节** | **文章与视频封面设计器**:根据全文或视频字幕自动提炼 200 字摘要与受众画像,匹配风格色彩,先隐喻后出图,视频封面自动注入吸睛网感指令。 |
  34. | **长文插图风格杂乱,手动排版繁琐** | **长文配图规划与自动回填**:编辑级视觉点位决策,全篇单画风单色调统一,支持一键批量生图并排版回填 Markdown。 |
  35. | **影楼摄影企划繁杂,AI 人像出图难控** | **专业摄影与时空企划编排**:支持情侣、婚纱、写真、全家福等 10+ 拍摄类型,首轮出齐 3 个高差异时空方案(带 8 宫格提示词代码块),一键复制自行出图或选定直出,拒绝泛 AI 感。 |
  36. ## 🎯 适用人群与场景
  37. - **自媒体创作者**:小红书爆款封面与图文、公众号题图、朋友圈精选海报
  38. - **公众号与深度长文作者**:长文核心摘要提炼、高点击封面设计、全篇风格统一的批量插图规划与 Markdown 自动回填
  39. - **摄影爱好者与人像创作者**:情侣写真、婚纱样片、亲子全家福、古风汉服等 10+ 题材的专业时空分镜编排与 8 宫格样片快速出图
  40. - **知识科普博主**:对比清单、架构金字塔、步骤流程等高传播信息长图
  41. - **条漫与故事创作者**:四格故事、多格情绪条漫、剧情分镜脚本、绘本插图
  42. - **视觉与品牌设计师**:快速输出高调性手绘草图、创意活动海报与插画原型
  43. ## 安装
  44. 直接把本仓库地址发给 Codex,让 Codex 自动完成安装:
  45. > **“帮我安装这个 Skill:https://github.com/yang0/handraw-style”**
  46. Codex 会自动拉取仓库并配置好全部风格与排版资源。
  47. 本项目聚焦手绘图片、封面、插图、IP、电商和摄影设计。视频制作、配音、分镜画廊和渲染已拆到独立的 `handraw-video` 项目,不再随图片 Skill 安装;视频封面设计作为静态图片能力保留。
  48. 其他功能的包内资源、运行工具与在线服务边界见[项目依赖说明](DEPENDENCIES.md)。提交前可运行`python scripts/audit_release.py`检查本机路径、常见凭据和视频文件是否混入。
  49. 本次自包含与隐私修正的范围、验证和历史提交边界见[发布检查记录](RELEASE_AUDIT.md)。
  50. ## 这个 Skill 怎么用
  51. ### 1. 智能推荐模式(零门槛:不指定编号直接出图)
  52. 如果你不想翻阅图鉴挑选编号,**直接告诉 Skill 你想画什么即可**!Skill 会自动理解主题的情绪、氛围与场景,**自动为你匹配最契合的手绘风格与经典主题色组合**(附带推荐理由),直接交付完整提示词:
  53. - 零参数直出:`帮我生成猫咪在窗台晒太阳的治愈插画提示词,你帮我选风格和主题色(可组合)`
  54. > 💡 智能推荐组合:风格 `#FA-018 (#018) · Minimal Deadpan Dialogue Cartoon` + 主题色 `C-26 · 柿子橙`(理由:画面充满温暖治愈与松弛感),直接输出中英双语提示词。
  55. - 补齐模式:若你仅指定了风格(如 `FB-006 / 041号风格,主题:秋天的第一杯奶茶`),Skill 会自动推荐最搭的主题色(如 `C-01 克莱因蓝`);若仅指定了主题色,Skill 会自动推荐最搭的画风。
  56. ### 2. 精准指定模式(自选手绘编号、图型与主题色)
  57. 1. 打开 [风格图鉴 (STYLES.md)](STYLES.md)、[排版图鉴 (LAYOUTS.md)](LAYOUTS.md) 或 [单色图鉴 (COLORS.md)](COLORS.md),浏览直观效果并挑选编号;
  58. 2. 记下喜欢的编号,例如风格 `FB-006`(原 `041`)、图型 `SC-001`,或主题色 `C-01`;
  59. 3. 输入指令给 Skill,例如:
  60. - 单独调用风格:`FB-006号风格(或 041号),主题:秋天的第一杯奶茶`
  61. - 组合排版与风格:`图型:SC-001,风格:FB-006,主题:秋天的第一杯奶茶`
  62. - 指定主题色:`图型:SC-001,风格:FB-006,主题色:C-01,主题:秋天的第一杯奶茶`
  63. 4. 得到带风格名称、版式定义与主题色限定的中英双语提示词;
  64. 5. 复制到 Midjourney、DALL-E 3、Stable Diffusion、Flux 等生图 AI 中直接出图。
  65. ### 3. AI 海报设计模式(结构化极简提示词,原生图文一体排版)
  66. 需要制作节日宣发、活动主视觉或社交媒体高调性海报?直接告诉 Skill 你的海报主题即可!
  67. - **输入指令**:`请帮我出海报提示词, 主题:秋分`
  68. - **智能协同**:Skill 会自动分析主题,从手绘图库中匹配最契合的画风(如 `#FE-053 (#268) · 当代人文水墨漫画`)与经典主题色(如 `主色 C-26 柿子橙 + 点缀色 C-03 普鲁士蓝`),按**8 大核心纯净设计字段**(主题、场景、受众、密度、情感、主题色、editorial风格、画风)输出提示词并附加图文一体指导规范。
  69. - **实测出图效果**:将生成的提示词直接输入生图模型(如 GPT Image 2/2.5),AI 模型将自主完成绝妙的排版构图与字画融合——大字书法标题、诗意对联、印章、日夜均分同辉意象与松弛水墨人物浑然一体:
  70. <p align="center">
  71. <img src="images/poster-demo-autumn-equinox.webp" width="360" alt="秋分节气海报实测生成效果">
  72. <br>
  73. <em>实测出图案例:#FE-053 (#268) 当代人文水墨漫画 + 主色 C-26 柿子橙 / 点缀色 C-03 普鲁士蓝</em>
  74. </p>
  75. <details>
  76. <summary>👉 点击展开查看完整海报提示词</summary>
  77. ```text
  78. 主题:秋分·昼夜均而寒暑平
  79. 业务场景:二十四节气传统文化海报、文创书店/茶饮节日宣发主视觉、社媒节气签
  80. 受众:传统文化爱好者、文艺青年、生活美学追求者、大众社媒读者
  81. 内容密度:低
  82. 情感基调:松弛闲适、温润拙朴、平衡从容、金秋诗意
  83. 主题色:主色为柿子橙(Persimmon Orange),点缀色为普鲁士蓝(Prussian Blue)
  84. editorial风格:Contemporary Literati Ink Cartoon Editorial 风格
  85. 画风:风格名称:#FE-053 (#268) · Contemporary Literati Ink Cartoon。参考作者/风格名称:当代人文水墨漫画。核心风格特征:现代生活人物、动物和日常小场景;白色或微暖纸面、大面积留白;毛笔和墨线快速画人物;线条松弛、粗细不一、枯湿自然、允许断笔和不闭合轮廓;人物高度概括、比例略笨拙,以姿态和动作表达情绪;人物是画面的主要视觉主体;场景保留少量必要元素;墨色为主,少量朱红、赭石、石绿、湖蓝局部点染;书法题字和朱红印章自然嵌入留白;整体拙朴、松弛、幽默、闲适;现代生活速写与中国写意笔墨结合。【如果主题直白包含画面元素那就按主题出图,文案由你来升华,但是不要直接描述画面。 如果主题比较概念化,那么文案和主题尽量保持一致,如果文案较长由你提炼,由你先设计画面隐喻(人类和非人类都行)再出图 。 文字参与构图,图文一体】
  86. ```
  87. </details>
  88. ### 4. 进阶心法:万能海报思维法(把信息图当高密度海报做)
  89. 觉得现有的信息图图型不够用?**不要让固定的表格和模版框住你的表达!**
  90. 信息图的本质是一张高信息密度的视觉海报。直接告诉 Skill 你的**分段时段、核心数据、步骤与使用场景**,生图模型将原生把时间轴、实景插图、手帐便签与文字融为一体:
  91. - **文旅出行路线**:《清迈 5 日慢游 DAY 1 导览》(分时段路线 + JJ市集/艺术村/宁曼路地标 + 泰北咖喱面与实用提醒)
  92. - **户外装备清单**:《山系露营轻量化装包指南》(睡眠/烹饪/照明三大系统平铺展开 + 8kg 控制,工装风)
  93. - **生活技能步骤**:《新手手冲咖啡指南》(水粉比 1:15 + 3 段注水时间轴 + 风味轮)
  94. 💡 **[👉 点击进入完整实战技巧与进阶教程 (TUTORIALS.md)](TUTORIALS.md)**(涵盖万能海报思维法、智能推荐机制、三元精准组装法、纯图/图文双模式与跨模型垫图策略等 5 大核心技巧与实战案例)。
  95. ### 5. 文章与视频封面设计模式(提炼摘要、受众推算、视觉隐喻与网感吸睛封面)
  96. 写完文章或剪完视频不知道怎么配一张高点击、有调性的封面?直接把文章或字幕文件发给 Skill 即可!
  97. - **适用场景**:微信公众号文章封面(首选 `2.35:1` 或 `21:9`)、X/Twitter 文章封面(`5:2`)、小红书封面(`3:4`)、视频封面/B站/YouTube/短视频封面(横屏 `16:9`,竖屏 `9:16` 或 `3:4`)。
  98. - **输入指令**:
  99. - 文章:`请根据这篇文章帮我设计封面:[粘贴文章全文或提供本地文章路径 d:\path\to\article.md]`
  100. - 视频/字幕:`请帮我设计视频封面:[粘贴字幕文本或输入 .srt/.vtt 字幕文件路径]`
  101. - **自动化工作流**:
  102. 1. **提炼核心摘要**:自动提炼 200 字左右的核心论点、剧情主线与认知启示;
  103. 2. **推算受众画像**:结合知识深度与诉求,输出精准读者/观众画像;
  104. 3. **注入风格与色彩**:从全库 365 种视觉风格与 36 种经典配色中匹配最契合的组合(如风格 `#FA-018 (#018) · Minimal Deadpan Dialogue Cartoon` + 主题色 `C-01 经典蓝`);
  105. 4. **原生隐喻与吸睛指令**:
  106. - 文章封面:注入核心指令——*“先设计隐喻再出图,主标题明显,小字少或者没有小字。其他你帮我设计。”* 让生图模型自主完成高质量版式与视觉隐喻构图;
  107. - 视频封面/字幕输入:强制注入——*“视频封面设计,网感强,吸引人眼球。先设计隐喻再出图,主标题明显,小字少或者没有小字。其他你帮我设计。”*;
  108. 5. **双轨交付引导**:
  109. - **【方式 A · 自主生图】**:复制输出的中英双语提示词,前往常用生图工具出图;
  110. - **【方式 B · 全自动生图】**:直接回复 **“全自动出图”**,Skill 即可自动调用生图工具直接生成高质量封面图片。
  111. ### 6. 文章批量插图与自动回填模式(长文配图规划、风格统一、自动排版)
  112. 长文干货如果全篇都是密密麻麻的文字,读者很容易疲劳流失;但手动到处找图不仅风格割裂,排版还费时费力。本模式提供从**“视觉脉络规划 → 全文风格统领 → 批量生图 → 自动排版回填”**的一站式解决方案:
  113. - **输入指令**:`请帮我为这篇文章规划插图并完成配图:[粘贴文章全文或提供本地文章路径 d:\path\to\article.md]`
  114. - **自动化工作流**:
  115. 1. **编辑级视觉点位决策(Visual Editorial Judgment)**:拒绝机械式段落塞图,深度理解文章论证逻辑与情绪流转,在最能产生视觉共鸣的关键拐点(开篇立意、抽象隐喻、概念解析、流程推演、收尾升华)规划 2–5 处核心配图点位;
  116. 2. **全篇单风格与单色调统一**:整篇文章固定锁定同一套手绘风格编号(如 `#FC-029 (#276) · 当代水墨`)与主题色(如 `C-01 · 克莱因蓝`),统一采用适合长文阅读流的 `4:3` 比例(支持自定义留白度:正常/适中/多),彻底杜绝全篇图片画风混乱的问题;
  117. 3. **标准结构化插图方案**:明确标注每张插图的【插入锚点】、【配图定位】、【图片类型】(editorial illustration / conceptual diagram / metaphorical illustration 等)、【视觉构思】与中英双语生图提示词;
  118. 4. **双轨交付闭环**:
  119. - **【方式 A · 自主生图回填】**:复制提示词自行出图后,把生成的图片拖入对话或发回本地路径,Skill 会自动把图片精准插入到文章对应段落后;
  120. - **【方式 B · 全自动生图插入】**:直接回复 **“全自动生图”**,Skill 将全自动批量生成整套插图,并调用内置自动化排版引擎精准插入原文章,一键输出排版完毕的 Markdown 定稿文件(如 `article_illustrated.md`)。
  121. ### 7. 专业摄影写真企划模式(情侣/婚纱/写真/全家福,3候选方案与8宫格极简出图)
  122. 告别千篇一律的普通游客照与泛 AI 感人像!无论是情侣、婚纱、个人写真、亲子全家福、古风汉服,还是职场肖像与旅行抓拍,系统按专业影楼样片与 Editorial 杂志级标准进行时空与视觉编排:
  123. - **适用场景**:情侣摄影、婚纱企划、个人写真、亲子全家福、古风汉服、新中式、复古港风、Cosplay、旅行抓拍、职场肖像等 10+ 拍摄类型。
  124. - **输入指令**:
  125. `摄影写真模式,请先出3个方案供我挑选,拍摄类型:情侣`
  126. *(亦可在网页画廊 [tutorials.html](skills/handdraw-style-prompter/gallery/tutorials.html) 中切换到「摄影」模式,点选拍摄类型标签;主题为选填项,不填由 AI 自由发挥)*
  127. - **自动化工作流与核心特色**:
  128. 1. **第一步绝不擅自选定**:严格推荐 3 个具有明显地域跨度与世界观差异的候选方案(2 个现代真实世界著名地标 + 1 个穿越/古代/未来/虚构作品方案);
  129. 2. **逐人造型与妆发定稿**:全方位规划人物关系描述、独立逐人服装穿搭、Editorial 发型师与化妆师定稿(女性默认影楼柔焦精修质感,保留真实皮肤质感与自然光影);
  130. 3. **独立代码块与 8 宫格关键词**:3 个方案分别完整封装在独立代码片段(` ```text `)中,末尾直接附加出图关键词 `出8宫格图片。`;
  131. 4. **极简出图双体验**:
  132. - **【一键复制自行出图】**:点击任意代码块右上角一键复制完整提示词,直接粘贴到 Midjourney、Flux 或其他生图模型自行出图;
  133. - **【回复 A/B/C 直接出图】**:在对话中直接回复 `A`、`B` 或 `C`,AI 即刻基于该方案执行 8 宫格图片生成,彻底跳过冗长繁琐的方案计划等待!
  134. ## 两种模式,随时切换
  135. 同一个编号和主题可以随时在两种模式之间切换,无需重新选择风格。
  136. - **纯图模式**:让主题决定画面内容,适合只需要插画、不需要画中文字的场景。示例:`纯图模式,FB-006 (041)号风格,主题:秋天的第一杯奶茶`。
  137. - **图文模式**:保留你写下的主题原文,让文字参与画面构图,适合金句、海报与社交媒体配图。示例:`图文模式,FE-052 (267)号风格,主题:有很多事你当时想不通,别着急,过段时间你可能就忘了。`。
  138. 切换时直接说“切换为纯图模式”或“切换为图文模式”即可;不指定时默认使用纯图模式。
  139. ### 图文模式演示
  140. 下图展示了文字与插图共同构图的效果;它仅作为图文模式演示,不绑定任何编号风格。
  141. ![图文模式演示:鸭子与手写文字](images/graphic-text-mode-demo.webp)
  142. ### 模型适配机制说明
  143. Skill 默认只负责把想法变成提示词;你明确要求“生图”时,会根据模型的已知能力决策最适合的生成策略:
  144. - **显式标定模型(首发深度适配 `gpt-image-2`)**:建立了全量风格的激活能力清单。生图优先级为:作者名称 + 风格名称 → 注入正向核心风格特征 → 仅在特征不足或难以激活时才传参考图,避免无谓传图导致主体构图被锁死。这个判断基于实际盲测,与作者是否著名无关。
  145. - **通用模型与未标定模型(如 Midjourney、Flux、Stable Diffusion、Imagen、Gemini 等)**:采用经过实测最稳妥的**“参考图高保真兜底(Reference Image Fallback)”**策略。系统会自动提供 4-grid 标准参考垫图或单图指引,确保即使在未深度标定关键词能力的模型中,也能 100% 忠实还原该编号对应的线条、质感与配色风格。
  146. - **欢迎社区共建**:能力矩阵定义在 `skills/handdraw-style-prompter/references/model_capabilities.json` 中,结构透明开放,欢迎社区提交更多生图模型的实测激活评测数据!
  147. ## 三个例子
  148. ```text
  149. FB-006 (041)号风格,主题:秋天的第一杯奶茶
  150. FE-041 (210)号风格,主题:小男孩在雪地里点鞭炮
  151. FH-039 (193)号风格,主题:大唐夜宴
  152. ```
  153. ## 风格精选预览
  154. 这里展示了手绘风格的精选预览拼图(FA-001–FA-016,原 001–016):
  155. ![风格精选预览(FA-001–FA-016)](images/A_001-016.webp)
  156. - 🖼️ **[👉 查看更多手绘与写实风格图片(370 种完整风格拼图大表,支持 FA~FI / 001–370 检索)](STYLES.md)**
  157. - 📄 **[查看风格详细参数表(370 种风格 Markdown 详细表格与特征描述)](styles_200_reorganized.md)**
  158. - 💻 *(本地离线使用:可在本地浏览器打开 `skills/handdraw-style-prompter/gallery/index.html` 交互检索与放大)*
  159. ## 排版图型展示
  160. 除了丰富的手绘风格,本库还收录了 **170 种排版图型**,涵盖社交卡片、数据长图、分镜故事、IP设计与电商,支持一键组合风格与排版出图。
  161. ### 1. 社媒卡(Social Cards · 21 种)
  162. 适合小红书、朋友圈、公众号配图及观点金句卡片,结构包含上下图文、文案主导、双格对照、签名展示等。
  163. ![社媒卡分类视口预览](images/layouts/preview-social-cards.webp)
  164. 👉 **[查看更多社媒卡图型(21 种完整图鉴与排版提示词)](LAYOUTS.md#social-cards)**
  165. ---
  166. ### 2. 信息图(Infographics · 35 种)
  167. 适合知识科普、对比清单、流程步骤及数据展示,结构包含金字塔层级、中心主图标注、多行多列对比等。
  168. ![信息图分类视口预览](images/layouts/preview-infographics.webp)
  169. 👉 **[查看更多信息图图型(35 种完整图鉴与排版提示词)](LAYOUTS.md#infographics)**
  170. ---
  171. ### 3. 漫画分镜(Comic Storyboards · 68 种)
  172. 适合多格叙事、条漫剧情、情绪递进及动态视觉,包含规则四格、起承转合、大格冲击、对角切割等专业分镜。
  173. ![漫画分镜分类视口预览](images/layouts/preview-comic-storyboards.webp)
  174. 👉 **[查看更多漫画分镜图型(68 种完整图鉴与排版提示词)](LAYOUTS.md#comic-storyboards)**
  175. ---
  176. ### 4. IP设计(IP Design · 13 种)
  177. 适合品牌 IP 策划、角色设定、情绪定位与商业衍生,打造高辨识度吉祥物与品牌形象封面。
  178. 👉 **[查看 IP 设计图型(13 种完整图鉴与排版提示词)](LAYOUTS.md#ip-characters)**
  179. ---
  180. ### 5. 电商(E-commerce · 24 种)
  181. 适合电商单品主图、多视角、局部细节特写、包装套装、卖点标注、参数尺寸、材质解析、选购决策、工艺说明、数据证明、口碑评价、品牌故事、新品上市、促销活动、内容种草及广告主视觉等高转化商业构图。
  182. 👉 **[查看电商图型(24 种完整图鉴与排版提示词)](LAYOUTS.md#ecommerce)**
  183. ---
  184. - 💡 **[👉 进入排版图型完整图鉴,浏览全部 170 种排版与提示词 ↗](LAYOUTS.md)**
  185. - 💻 *(本地离线使用:可在本地浏览器打开 `skills/handdraw-style-prompter/gallery/layouts.html` 交互检索与放大)*
  186. ---
  187. ## 经典单色主题色展示(36 种)
  188. 本项目精选并提炼了 **36 种经典单色主题色**(编号 `C-01` ~ `C-36`,涵盖经典蓝系、清新绿系、古典红绿、浪漫粉紫、暖阳大地、中性色系 6 大色系)。
  189. 无论搭配何种手绘风格与排版结构,只需指定主题色编号,即可统领整张画面的色彩基调;也可在画廊中直接一键复制单色提示词。
  190. ![经典单色库预览](images/colors/sheet_01.webp)
  191. - 🎨 **[👉 进入经典单色主题色完整图鉴,浏览全部 36 种单色与提示词 ↗](COLORS.md)**
  192. - 💻 *(本地离线使用:可在本地浏览器打开 `skills/handdraw-style-prompter/gallery/colors.html` 一键复制中英提示词与交互预览)*
  193. ---
  194. ## 作者与社群交流
  195. - 📚 **官方飞书知识库**:[点击访问知识库(速查画廊 / 常见问题速查 / 创业变现案例) ↗](https://ecmn3m3i17.feishu.cn/wiki/space/7689222460876524523?ccm_open_type=lark_wiki_spaceLink&open_tab_from=wiki_home)
  196. - **创作变现交流群 / 作者微信**:请优先加群,满了的话也可以尝试加我个人微信。请备注:**手绘**
  197. | ① 优先加入群聊(手绘12群) | ② 个人微信(备用) |
  198. | :---: | :---: |
  199. | <img src="images/wechat_group.png" alt="微信群聊二维码" width="240"> | <img src="images/wechat_personal.png" alt="个人微信二维码" width="240"> |
  200. - X:[@yang02010](https://x.com/yang02010)
  201. ---
  202. ## 📄 开源许可协议 (License)
  203. 本项目采用宽松且附带署名要求的开源协议。**欢迎自由使用、修改和商业化。**
  204. 唯一的小请求:如果这个项目对你有帮助,或者你的项目基于它进行了二次开发,请保留作者 **yang0** 和原仓库地址:[https://github.com/yang0/handraw-style](https://github.com/yang0/handraw-style)。
  205. 完整许可协议详见 [LICENSE](LICENSE)。

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.