VIS STUDIO

English | 中文

English

VIS STUDIO is an AI-native operating system for brand identity and visual identity system work.

It helps teams move from messy inputs to a reusable brand system that AI can actually follow: strategy, visual direction, identity system, applications, governance, wrappers, adapters, and brand-specific prompts.

This is not just a prompt pack. This is not just a logo workflow. This is not just a UI DESIGN.md.

VIS STUDIO is built for full VIS work.

What VIS STUDIO is

VIS STUDIO turns brand work into a structured production system:

• discovery and strategic diagnosis
• reference distillation instead of reference copying
• concept exploration and territory narrowing
• identity-system production
• application-system planning and routing
• governance handoff and downstream reuse

It packages the result into machine-readable outputs so AI tools can work inside the approved system instead of improvising from vague style language.

Why it exists

Most AI design workflows break in the same place:

• strategy is missing
• references are copied instead of distilled
• application work is treated as an afterthought
• tools are prompted inconsistently
• every new asset restarts from zero

VIS STUDIO solves that by making the brand system explicit, reusable, and automation-ready.

What it produces

Core design documents:

• REFERENCE_STYLE_DISTILLATION.md
• BRAND_FOUNDATION_DESIGN.md
• UI_UX_DESIGN.md
• APPLICATION_DESIGN.md
• DESIGN_INDEX.md

Machine-readable packaging:

• brand_pack.json

Reusable downstream wrappers:

• <brand-slug>-brand-applications
• <brand-slug>-brand-guidelines

Tool-enablement layer:

• TOOL_ADAPTER_INDEX.md
• references/tool-adapters/*.md
• TOOL_PROMPT_SKELETON_INDEX.md
• references/tool-prompt-skeletons/*.md

What gets generated automatically

When a project reaches the packaging stage, VIS STUDIO can automatically generate:

• the final brand_pack.json
• a brand-specific applications wrapper
• a brand-specific guidelines wrapper
• bundled tool adapters inside the generated wrapper
• brand-specific prompt files inside the generated wrapper

Those brand-specific prompt files currently support:

• Figma MCP
• Stitch
• Pencil
• Adobe Illustrator
• Adobe Photoshop
• Canva
• Inkscape
• image-generation models such as Doubao, Kling, Nano Banana, and similar tools

Workflow

1. brand-discovery-strategy Diagnose the route, check input completeness, ask targeted questions when critical information is missing, and build the strategic base.
2. visual-concept-exploration Distill references, translate strategy into visual territories, and narrow direction.
3. identity-system-production Build the identity core and first structured brand payload.
4. brand-application-system Convert the identity into application families, routes, and production logic.
5. brand-governance-rollout Finalize governance, packaging, and downstream reuse.

What makes it different

• It treats VIS as a system, not a one-off design exercise.
• It separates identity production from application production.
• It forces reference distillation before visual execution.
• It supports targeted intake questions when inputs are incomplete.
• It creates outputs that downstream AI tools can reliably consume.
• It includes both generic tool skeletons and brand-specific prompt generation.

Methodology inside the system

VIS STUDIO also incorporates useful methods from the local visual-identity-direction skill, including:

• strategy-to-visual translation
• creative-brief discipline
• moodboard rationale
• logo-brief structure
• photography direction
• typography and color reasoning

Public skills

• brand-discovery-strategy
• visual-concept-exploration
• identity-system-production
• brand-application-system
• brand-governance-rollout

Support skills

• brand-style-playbook-selector
• application-scope-planner
• application-route-classifier
• application-template-factory
• application-mockup-composer
• brand-application-factory

Repository structure

• skills/ Public VIS workflow skills
• support/ Internal support skills for routing, packaging, and production logic
• references/ Shared architecture notes, adapters, and prompt system references
• README.md Product overview
• CHANGELOG.md Release history

Important notes

• brand_pack.json is the downstream source of truth.
• Illustrator-heavy work is still treated as manual_vector.
• Canva-friendly work follows template logic.
• Photoshop-friendly work follows mockup logic.
• The wrapper layer is intentionally thin: it applies the approved system, not redefines it.

中文

VIS STUDIO 是一套面向品牌识别与视觉识别系统工作的 AI 原生操作系统。

它的目标，是把零散、模糊、依赖设计师个人脑内判断的品牌项目，转成 AI 也能稳定执行的结构化系统：从策略、方向、识别系统、应用扩展、治理交付，到 wrapper、adapter 和品牌专属 prompt，全部纳入同一条生产链。

它不是普通的 prompt 包。 它不是单独的 logo 工作流。 它也不只是一个面向 UI 的 DESIGN.md 套件。

VIS STUDIO 面向的是完整 VIS 工作。

VIS STUDIO 是什么

VIS STUDIO 把品牌工作变成一套可执行系统：

• discovery 与策略诊断
• 参考风格提炼，而不是照抄参考
• 概念探索与方向收敛
• 识别系统生产
• 应用系统规划与路由
• 治理交接与后续复用

它会把结果封装成机器可读输出，让 AI 工具不再依赖模糊的审美描述，而是依据已批准的品牌系统工作。

为什么要做它

大多数 AI 设计流程的问题都出在同几个地方：

• 没有策略基础
• 参考没有提炼，只有拼贴模仿
• application 被当成附属工作
• 不同工具的调用方式不统一
• 每做一个新物料都像重新开局

VIS STUDIO 的作用，就是把品牌系统显式化、可复用化、自动化。

它会产出什么

核心设计文件：

• REFERENCE_STYLE_DISTILLATION.md
• BRAND_FOUNDATION_DESIGN.md
• UI_UX_DESIGN.md
• APPLICATION_DESIGN.md
• DESIGN_INDEX.md

机器可读打包层：

• brand_pack.json

可复用下游 wrapper：

• <brand-slug>-brand-applications
• <brand-slug>-brand-guidelines

工具协同层：

• TOOL_ADAPTER_INDEX.md
• references/tool-adapters/*.md
• TOOL_PROMPT_SKELETON_INDEX.md
• references/tool-prompt-skeletons/*.md

哪些内容会自动生成

当项目走到 packaging 阶段，VIS STUDIO 可以自动生成：

• 最终 brand_pack.json
• 品牌专属 applications wrapper
• 品牌专属 guidelines wrapper
• 每个 wrapper 内自带的一套 tool adapters
• 每个 wrapper 内自带的一套品牌专属 prompt 文件

这些品牌专属 prompt 目前支持：

• Figma MCP
• Stitch
• Pencil
• Adobe Illustrator
• Adobe Photoshop
• Canva
• Inkscape
• 文生图模型，如豆包、可灵、Nano Banana 及类似工具

工作流

1. brand-discovery-strategy 先判断项目路径，检查输入是否完整；如果关键信息缺失，就先追问，再建立策略基础。
2. visual-concept-exploration 提炼参考风格，把策略翻译成视觉方向，并收敛方案。
3. identity-system-production 构建识别系统核心，并产出第一版结构化品牌数据。
4. brand-application-system 把识别系统扩展成应用家族、生产路线和规模化逻辑。
5. brand-governance-rollout 完成治理、打包和后续复用层。

它和普通品牌技能的区别

• 它把 VIS 当成系统，而不是一次性设计任务。
• 它明确区分 identity production 和 application production。
• 它要求先做 reference distillation，再做视觉执行。
• 它支持缺信息时的 targeted intake questions。
• 它会产出 AI 工具可以稳定消费的结构化结果。
• 它同时提供通用 skeleton 和品牌专属 prompt 自动生成。

系统内部的方法论

VIS STUDIO 同时吸收了本地 visual-identity-direction skill 中有价值的方法，包括：

• strategy-to-visual translation
• creative brief discipline
• moodboard rationale
• logo brief structure
• photography direction
• typography and color reasoning

公开技能

• brand-discovery-strategy
• visual-concept-exploration
• identity-system-production
• brand-application-system
• brand-governance-rollout

支持技能

• brand-style-playbook-selector
• application-scope-planner
• application-route-classifier
• application-template-factory
• application-mockup-composer
• brand-application-factory

仓库结构

• skills/ 公开 VIS 工作流技能
• support/ 内部 routing、packaging、production support 技能
• references/ 共享架构说明、adapter 和 prompt 系统参考
• README.md 产品化总览
• CHANGELOG.md 发布历史

重要说明

• brand_pack.json 是后续自动化的事实来源。
• Illustrator 重型工作仍然归为 manual_vector。
• 适合 Canva 的内容走模板路线。
• 适合 Photoshop 的内容走样机路线。
• wrapper 层刻意保持轻量，它负责应用已批准系统，而不是重新定义系统。