image-assistant
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chinese配图助手
Image Assistant
触发方式
Trigger Methods
当用户说类似以下内容时触发：
- “这段内容做个图 / 配几张图？”
- “给我两张（或多张）出图提示词”
- “字太多不好看，帮我更趣味、更好读”
- “把这个流程封装成提示词模板/skills”
- “/image “/配图” “/出图”
Trigger when the user says something like:
- "Make a graphic for this content / create several supporting images?"
- "Give me two (or multiple) image generation prompts"
- "Too much text, make it more interesting and readable"
- "Package this process into a prompt template/skills"
- "/image " "/配图" "/出图"
流程概览
Process Overview
| 阶段 | 名称 | 目标 | 详细文件 |
|---|---|---|---|
| 1 | 需求澄清（Spec/DoD） | 先挖需求：内容/场景/受众/字多字少；尺寸与风格默认；产出一句话复述与需求小结 | |
| 2 | 配图规划（拆块→清单） | 拆内容→定图清单（几张/每张讲啥/用啥模板） | |
| 3 | 文案定稿（Copy Spec） | 逐字定稿“图上写什么”（唯一真值） | |
| 4 | 提示词封装（Prompt Pack） | 把 Copy Spec 封装成可复制提示词；生成批量请求包并在用户确认后批量出图 | |
| 5 | 迭代润色 | 根据反馈减字、换隐喻、提可读性 | |
| Stage | Name | Goal | Detailed File |
|---|---|---|---|
| 1 | Requirements Clarification (Spec/DoD) | First explore requirements: content/scenario/audience/text density; default size and style; produce a one-sentence restatement and requirement summary | |
| 2 | Image Planning (Chunking → List) | Chunk content → define image list (quantity / what each image covers / recommended template) | |
| 3 | Copy Finalization (Copy Spec) | Finalize "what text to put on the image" word by word (single source of truth) | |
| 4 | Prompt Packaging (Prompt Pack) | Package Copy Spec into copyable prompts; generate batch request packages and batch generate images after user confirmation | |
| 5 | Iterative Polishing | Reduce text, replace metaphors, and improve readability based on feedback | |
调度规则
Scheduling Rules
如何判断当前阶段：
- 还没把需求讲清楚（内容 + 场景 + 受众 + 字多/字少）→ 阶段1
- 文章很长、需要拆块，或需要确定“几张图/每张讲什么”→ 阶段2
- 已确认图清单，但还没确定“图上逐字写什么”→ 阶段3
- Copy Spec 已确认，要出可复制提示词；（可选）提示词确认后批量出图 → 阶段4
- 用户反馈“字多/不好看/不符合封面” → 阶段5（必要时回退到阶段1重锁需求与字多/字少）
每个阶段开始时：
- 告诉用户当前阶段与本阶段输出物
- 读取对应阶段文件并按步骤执行
How to determine the current stage:
- Requirements not clearly defined (content + scenario + audience + text density) → Stage 1
- Long article that needs chunking, or need to confirm "how many images / what each image covers" → Stage 2
- Image list confirmed, but "exact text for each image" not finalized → Stage 3
- Copy Spec confirmed, need to generate copyable prompts; (optional) batch generate images after prompt confirmation → Stage 4
- User feedbacks "too much text / not good-looking / doesn't fit the cover" → Stage 5 (roll back to Stage 1 to re-lock requirements and text density if necessary)
At the start of each stage:
- Inform the user of the current stage and deliverables for this stage
- Read the corresponding stage file and execute according to steps
输出规范（必须遵守）
Output Specifications (Must Follow)
- 每张图一个“核心信息”，不把解释性段落塞进图里
- 所有中文必须清晰可读：大字号、少字短句、避免密集小字
- 每张提示词用一个独立代码块输出，便于复制
- 默认输出 16:9 横版（除非用户明确要 3:4 漫画/竖版）
- 默认风格：奶油纸底 + 彩铅水彩手绘 + 轻涂鸦，趣味但干净（可用 ）
templates/style-block.md - 阶段3产物（Copy Spec）一旦确认，阶段4不得擅自改文案，只做封装与参数/约束补全
- Each image has one "core message", do not include explanatory paragraphs in the image
- All Chinese text must be clearly readable: large font size, short sentences with minimal text, avoid dense small text
- Each prompt is output in an independent code block for easy copying
- Default output is 16:9 landscape (unless the user explicitly requests 3:4 comic/portrait)
- Default style: cream paper background + colored pencil watercolor hand-drawn + light doodles, fun but clean (can use )
templates/style-block.md - Once the Stage 3 deliverable (Copy Spec) is confirmed, the copy cannot be modified without permission in Stage 4; only packaging and parameter/constraint completion are allowed
快速使用（给用户的最小输入）
Quick Start (Minimum Input for Users)
用户只要给这四项，就能开始：
- 要配图的内容（可是一段、一个小节、或整篇文章）
- 用在哪里 + 观看距离（PPT投影远看 / 手机近看 / 海报）
- 谁来看（小白/从业者/老板/学生…）
- 偏好：更“少字清爽”还是更“信息密度”
可选补充（不写也没关系）：
- 你大概想要哪类图：封面/目录、单页概览、讲义解释、社媒海报（不确定我会根据场景与偏好推荐）
你要做的交付顺序：
- 先输出：图清单（几张 + 每张一句话目的 + 模板建议）（阶段2）
- 用户确认后：逐张输出 Copy Spec（逐字定稿）（阶段3）
- Copy Spec 确认后：逐张输出可复制提示词/调用包（阶段4）
- （可选）提示词确认后：批量调用 API 出图（阶段4内连续流程）
- 用户说“字多/不好看”就进入迭代（阶段5）
Users only need to provide these four items to start:
- Content to be illustrated (can be a paragraph, a section, or an entire article)
- Usage scenario + viewing distance (PPT projection for long-distance viewing / mobile phone for close viewing / poster)
- Target audience (beginners/practitioners/managers/students...)
- Preference: "more minimal and clean" or "higher information density"
Optional supplements (not required):
- What type of images you roughly want: cover/table of contents, single-page overview, lecture explanation, social media poster (I will recommend based on scenario and preference if unsure)
Your delivery order:
- First output: Image list (quantity + one-sentence purpose for each + template recommendation) (Stage 2)
- After user confirmation: Output Copy Spec for each image (finalize text word by word) (Stage 3)
- After Copy Spec confirmation: Output copyable prompts/call packages for each image (Stage 4)
- (Optional) After prompt confirmation: Batch call API to generate images (continuous process within Stage 4)
- Enter iteration when the user says "too much text / not good-looking" (Stage 5)
文件结构
File Structure
stages/
├── 01-brief.md
├── 02-plan.md
├── 03-copy.md
├── 04-prompts.md
└── 05-iterate.md
templates/
├── style-block.md
├── 16x9-infographic.md
├── 16x9-contrast-2cards.md
├── 16x9-3cards-insights.md
├── 16x9-cover-roadmap.md
├── 16x9-5panel-comic.md
├── api-config.md
├── apimart-curl.md
├── apimart-requests-jsonl.md
└── checklist.md
examples/
└── ai-tools-selection.md
scripts/
├── apimart_batch_generate.py
├── apimart.env.example
└── README.mdstages/
├── 01-brief.md
├── 02-plan.md
├── 03-copy.md
├── 04-prompts.md
└── 05-iterate.md
templates/
├── style-block.md
├── 16x9-infographic.md
├── 16x9-contrast-2cards.md
├── 16x9-3cards-insights.md
├── 16x9-cover-roadmap.md
├── 16x9-5panel-comic.md
├── api-config.md
├── apimart-curl.md
├── apimart-requests-jsonl.md
└── checklist.md
examples/
└── ai-tools-selection.md
scripts/
├── apimart_batch_generate.py
├── apimart.env.example
└── README.md