prism

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

PRISM — People Research & Insight System for Makers

PRISM — 面向创作者的用户研究与洞察系统

Two modules. One skill. From research insight to design specification.

一个技能，两大模块。从研究洞察到设计规范，一站式完成。

Module 1: UX Research

模块1：UX研究

Generate research-grade deliverables — personas, journey maps, interview synthesis, and usability test plans. Every output should be something a senior UX researcher would be comfortable presenting to stakeholders.

生成专业级研究交付物——用户画像、客户旅程地图、访谈综合报告和可用性测试计划。所有输出内容都达到资深UX研究员可向利益相关方展示的水准。

Module 2: UI Design

模块2：UI设计

Generate professional design deliverables — design philosophies, component specs, design tokens, layout systems, interaction specs, and visual canvas artifacts. Every output should look like it came from a senior product designer who obsesses over craft.

生成专业设计交付物——设计理念、组件规范、设计令牌、布局系统、交互规范和视觉画布制品。所有输出内容都如同出自对工艺精益求精的资深产品设计师之手。

Pipeline Overview

流程概述

User Input → Identify Module → Identify Deliverable → Classify Input → Generate → Quality Check → Output

用户输入 → 识别模块 → 识别交付物 → 分类输入 → 生成内容 → 质量检查 → 输出结果

Step 1: Identify the Module & Deliverable

步骤1：识别模块与交付物

UX Research Deliverables

UX研究交付物

If the user wants to understand...	Generate
Who their users are	Personas — see `references/PERSONA-GUIDE.md`
What users experience across touchpoints	Journey Map — see `references/JOURNEY-MAP-GUIDE.md`
Patterns in raw research data	Interview Synthesis — see `references/SYNTHESIS-GUIDE.md`
How to test an experience	Usability Test Plan — see `references/TEST-PLAN-GUIDE.md`

如果用户想要了解...	生成内容
用户是谁	用户画像（Personas） — 参考 `references/PERSONA-GUIDE.md`
用户在各个触点的体验	客户旅程地图（Journey Map） — 参考 `references/JOURNEY-MAP-GUIDE.md`
原始研究数据中的规律	访谈综合报告（Interview Synthesis） — 参考 `references/SYNTHESIS-GUIDE.md`
如何测试体验	可用性测试计划（Usability Test Plan） — 参考 `references/TEST-PLAN-GUIDE.md`

UI Design Deliverables

UI设计交付物

If the user wants to define...	Generate
Visual direction for a product	Design Philosophy — see `references/PHILOSOPHY-GUIDE.md`
Component patterns for an interface	Component Spec — see `references/COMPONENT-GUIDE.md`
Color, type, spacing foundations	Design Tokens — see `references/TOKEN-GUIDE.md`
Page structure & layout logic	Layout System — see `references/LAYOUT-GUIDE.md`
Interaction & motion behavior	Interaction Spec — see `references/INTERACTION-GUIDE.md`
Visual artifact (poster, infographic, PDF)	Canvas Artifact — see `references/CANVAS-GUIDE.md`

If the request is ambiguous, ask the user.

Research order (each builds on the previous): Interview Synthesis → Personas → Journey Maps → Usability Test Plans.

Design order (each builds on the previous): Design Philosophy → Design Tokens → Layout System → Component Spec → Interaction Spec.

Cross-module flow: Research deliverables feed design deliverables. Personas → emotional targets. Journey maps → key moments for state design. Synthesis → pain point to design solution mapping.

如果用户想要定义...	生成内容
产品的视觉方向	设计理念（Design Philosophy） — 参考 `references/PHILOSOPHY-GUIDE.md`
界面的组件模式	组件规范（Component Spec） — 参考 `references/COMPONENT-GUIDE.md`
颜色、字体、间距基础	设计令牌（Design Tokens） — 参考 `references/TOKEN-GUIDE.md`
页面结构与布局逻辑	布局系统（Layout System） — 参考 `references/LAYOUT-GUIDE.md`
交互与动效行为	交互规范（Interaction Spec） — 参考 `references/INTERACTION-GUIDE.md`
视觉制品（海报、信息图、PDF）	画布制品（Canvas Artifact） — 参考 `references/CANVAS-GUIDE.md`

如果请求模糊不清，请询问用户。

研究顺序（每项基于前一项构建）：访谈综合报告 → 用户画像 → 客户旅程地图 → 可用性测试计划。

设计顺序（每项基于前一项构建）：设计理念 → 设计令牌 → 布局系统 → 组件规范 → 交互规范。

跨模块流程：研究交付物为设计交付物提供依据。用户画像 → 情感目标。客户旅程地图 → 状态设计的关键节点。访谈综合报告 → 痛点到设计解决方案的映射。

Step 2: Classify Input Fidelity

步骤2：分类输入保真度

Input type	Approach
Raw interview notes / transcripts	Synthesize — find patterns, surface insights, use direct quotes. Never invent data.
Survey data	Identify quantitative trends, pull qualitative verbatims, note sample sizes.
Analytics / metrics	Ground personas and journeys in real usage patterns.
Existing brand guidelines / style guide	Extend — build on what exists, don't contradict it.
Product screenshots / competitor references	Analyze — extract patterns, identify gaps, propose improvements.
PRISM research artifacts	Ground — connect every design choice to a user need.
Brief / description only	Generate a well-reasoned draft, but label every assumption: "Assumption — validate with research."

输入类型	处理方式
原始访谈笔记/转录稿	综合分析——发现规律、提炼洞察、使用直接引语。绝不编造数据。
调研数据	识别量化趋势、提取定性原话、标注样本量。
分析数据/指标	基于真实使用模式构建用户画像和旅程地图。
现有品牌指南/风格指南	拓展延伸——在现有基础上构建，不与原有内容冲突。
产品截图/竞品参考	分析——提取模式、识别差距、提出改进方案。
PRISM研究制品	锚定核心——将每个设计决策与用户需求关联。
仅简报/描述	生成合理的草稿，但需明确标注所有假设：“假设——需通过研究验证。”

Step 3: Generate the Deliverable

步骤3：生成交付物

Read the relevant reference guide and follow its template exactly. Each guide contains the methodology, the output template, guidelines, and examples.

For UI Design deliverables: Always establish a design philosophy first (see

references/PHILOSOPHY-GUIDE.md

). Every design system needs a soul before it has rules.

阅读相关参考指南并严格遵循其模板。每个指南包含方法论、输出模板、准则和示例。

对于UI设计交付物：始终先确立设计理念（参考

references/PHILOSOPHY-GUIDE.md

）。每个设计系统在有规则之前，先要有灵魂。

Step 4: Financial Services Layer (Optional)

步骤4：金融服务层（可选）

Activate when the user mentions financial products, banking, payments, insurance, wealth management, lending, or regulatory compliance — or when they explicitly ask for it.

For Research deliverables: Read
```
references/FINSERV-LAYER.md
```
For Design deliverables: Read
```
references/FINSERV-UI-LAYER.md
```

当用户提及金融产品、银行、支付、保险、财富管理、借贷或监管合规，或明确要求时激活该层。

对于研究交付物：阅读
```
references/FINSERV-LAYER.md
```
对于设计交付物：阅读
```
references/FINSERV-UI-LAYER.md
```

Step 5: Quality Check

步骤5：质量检查

Before delivering, verify against

references/QUALITY-CHECKLIST.md

. Every deliverable must pass all checks.

交付前，对照

references/QUALITY-CHECKLIST.md

进行验证。所有交付物必须通过全部检查项。

Step 6: Output

步骤6：输出

Save all deliverables as Markdown (.md) files:

Research outputs:

```
personas.md
```
```
journey-map-[persona-name].md
```
```
interview-synthesis.md
```
```
usability-test-plan.md
```

Design outputs:

```
design-philosophy.md
```
```
design-tokens.md
```
```
component-spec-[name].md
```
```
layout-system.md
```
```
interaction-spec.md
```

When visual canvas artifacts are requested, also output as

.pdf

.png

将所有交付物保存为Markdown（.md）文件：

研究输出：

```
personas.md
```
```
journey-map-[persona-name].md
```
```
interview-synthesis.md
```
```
usability-test-plan.md
```

设计输出：

```
design-philosophy.md
```
```
design-tokens.md
```
```
component-spec-[name].md
```
```
layout-system.md
```
```
interaction-spec.md
```

当请求视觉画布制品时，同时输出为

.pdf

或

.png

格式。

Key Principles

核心原则

These apply to every deliverable this skill generates:

Research over opinion. When real data exists, use it. When it doesn't, say so clearly.
Synthesis over summary. Find patterns across data points — don't recap individual interviews.
Actionable over descriptive. Every section should connect to a design decision or recommendation.
Specific over generic. If a persona, journey, finding, or component could apply to any product, it's too vague.
Transparent over polished. Label assumptions, flag gaps, rate evidence strength. Stakeholders need to know what's proven and what's hypothesized.
Philosophy before pixels. Every design system needs a reason to exist. Establish the "why" before the "what."
Craft over templates. Design output must look meticulously crafted — painstaking attention, master-level execution.
Accessible by default. WCAG 2.2 AA minimum. Built into every spec, not bolted on after.
Buildable. Design specs must be precise enough that a developer can implement without guessing.

这些原则适用于该技能生成的所有交付物：

研究优先于主观意见。有真实数据时使用真实数据；没有时明确说明。
综合分析优先于简单总结。在数据点中寻找规律——不要复述单个访谈内容。
可行动优先于描述性。每个部分都应关联到设计决策或建议。
具体明确优先于泛泛而谈。如果用户画像、旅程、发现或组件适用于任何产品，则过于模糊。
透明公开优先于表面完美。标注假设、指出差距、评估证据强度。利益相关方需要知道哪些已被证实，哪些是假设。
理念优先于像素。每个设计系统都需要存在的理由。先确立“为什么”，再考虑“是什么”。
工艺优先于模板。设计输出必须看起来经过精心雕琢——细致入微的关注、大师级的执行。
默认无障碍。最低符合WCAG 2.2 AA标准。融入每个规范，而非事后添加。
可落地实现。设计规范必须足够精确，让开发者无需猜测即可实现。