poster-prompt-generator

Poster Prompt Generator

Core Design Philosophy: High-dimensional structured anchoring, deep collaboration with style and color galleries

[!IMPORTANT] [Core Principle: Fully dynamic aesthetic decision-making, strictly prohibited from hardcoding/mechanically applying]

• AI dynamically assesses in real time every time: Strictly prohibited from maintaining or applying any static rigid mapping of "theme keyword -> fixed style number/color number" (for example, strictly prohibited from recommending 268 whenever solar terms/tradition appear, recommending 266 whenever nature appears, recommending 011/054 whenever technology appears, and recommending 018 whenever life appears).
• Open-ended matching across the entire library: The pairing of poster style and color is absolutely not table lookup! Each time, the AI must, based on the deep context, spiritual core, visual metaphor, and pictorial composition of the current poster theme, conduct among the entire library of 278 hand-drawn styles (#001-#278) and 30 classic theme colors (C-01-C-30)real-time, open-ended aesthetic reasoning and dynamic pairing。
• Stimulate diversity across the entire library: The same theme has multidimensional aesthetic possibilities under different design perspectives (for example, "vegetable garden" can be ink wash freehand, pastoral woodcut printmaking, coarse oil pastel, or Bauhaus geometric composition). Each recommendation should be conceived according to the specific entry point, fully demonstrating the rich vitality of all 278 art styles and 30 colors in the library.
• Make aesthetic reasons concrete: The aesthetic recommendation reasons given by the AI must be closely tied to the visual metaphor and pictorial composition of the current poster theme, explaining why the brushwork/texture of that art style and the emotion of that color can perfectly convey this theme, and eliminating boilerplate.

The new generation of image generation models (such as GPT Image 2/2.5), after understanding the business scenario, audience, channel, and aesthetic style, possess extremely strong native autonomous composition capabilities.Poster prompts do not need lengthy discussion describing object details, camera positions, and light spots; they only need to precisely provide core business and design fields：

1. [Theme]: The core concept or main copy expressed by the poster;
2. [Business Scenario]: The specific placement or usage scenario;
3. [Audience]: Target audience group;
4. [Content Density]: Fill in only the level: very low / low / medium / high;
5. [Emotional Tone]: Core emotion and atmosphere keywords;
6. [Theme Color]: Recommend or specify from the project's 30 classic monochrome colors (C-01 ~ C-30). The number of colors is not hardcoded and is flexibly determined according to the visual hierarchy: monochrome dominance (such as C-01 克莱因蓝), two-color pairing (such as 主色 C-01 克莱因蓝 + 点缀色 C-28 暖杏仁) or three-color harmony;
7. [editorial style]: Strictly adopt the standard naming format to define the poster editorial layout and visual design tone, such as:Clean Airy Flat Vector Illustration Editorial 风格；
8. [Art Style]: Inject gallery hand-drawn style completely according to the project's hand-drawn library classic standard format:
   • Style name:#{编号} · {generation_name}
   • Reference author/style name:{reference}
   • Core style features and reference images (Strictly determine dynamically according to the model capability matrix）：
     • Strong author/strong name activation style(such as #042 Beatrix Potter、#011 David Shrigley、#002 Saul Steinberg , etc.): the model already has a deep semantic prior for the author name,when generating images, simply referencing the author is sufficient; outputting core style features is strictly prohibited, and outputting reference images is also strictly prohibited. The output format is:风格名称：#{编号} · {generation_name}。参考作者/风格名称：{reference}。
     • Strong feature activation style(such as #276、#277、#268 , etc.): output 核心风格特征：{traits}。, outputting reference images is strictly prohibited.
     • Weak activation/requires reference image style(such as #255 , etc.): output 核心风格特征：{traits}。 and dynamically append the reference image path and isolation statement.
9. [Default rule for integrated poster text and image]: poster design is essentially highly dependent on the topological composition of text and imagery.When making a poster, regardless of whether the user explicitly enables text-image mode, by default the text-image mode suffix must be unconditionally appended at the end of the prompt： 【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】

Fields within the prompt must strictly not have redundant descriptive embellishments or boilerplate added

Delivery specification: gallery recommended card + 8 pure design fields + default integrated text-image suffix

When receiving a poster design request,first output the gallery intelligent recommendation combination and aesthetic rationale above the prompt, then output the pure fields in the following format：

💡 Gallery intelligent recommendation combination：

• Hand-drawn style:#{number} · {Chinese style name} ({English Generation Name})
• Activation decision:{activation source and determination of no reference image needed/reference image needed/core features exempt}
• Classic theme colors:{freely decide on a single-color, two-color, or three-color combination, indicating the color number and name and role (such as primary color/accent color)}
• Aesthetic rationale:[1 sentence explaining why this art style and color combination fit the poster theme] (If you need to switch to another style or color from the gallery, you can tell me at any time)

主题：[核心主题词或标语]
业务场景：[投放场景]
受众：[目标人群]
内容密度：[极低 / 低 / 中 / 高]
情感基调：[核心情绪关键词]
主题色：[单色 / 双色（主色+点缀色） / 三色搭配]
editorial风格：[标准名称 Editorial 风格]
画风：风格名称：#{编号} · {generation_name}。参考作者/风格名称：{reference}。[仅当非强作者激活且需特征时输出：核心风格特征：{traits}。][若模型能力需垫图则动态追加参考图声明]【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】