agentic-ux-design---relationship-centric-interfaces by bencium/bencium-claude-code-design-skill
npx skills add https://github.com/bencium/bencium-claude-code-design-skill --skill 'Agentic UX Design - Relationship-Centric Interfaces'从以屏幕为中心到以关系为中心的设计范式转变。
传统的用户体验设计优化的是单个屏幕和孤立的交互。而 Agentic UX 设计的是持续的关系,在这种关系中,系统能够跨会话、跨设备、跨情境地学习、记忆并随着用户一同演进。
核心原则: 每一次交互都建立在已学习的偏好和用户历史之上。系统不仅仅是响应——它们会发展出随着时间推移而不断加深的理解。
在开始时声明: "我正在使用关系设计技能来创建一个具备代理性、记忆感知的界面,该界面能与用户建立长期关系。"
在以下情况下使用此技能:
何时不使用:
旧模式: 存储静态偏好(主题:深色,语言:英文)
新模式: 维护动态、演进的关系模型
设计目标:
广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
关键问题: 如果这个体验能记住一切并随时间推移而变得更好,它会是什么样子?
设计通过渐进式信任获得自主权的界面:
第一阶段:透明阶段
第二阶段:选择性披露阶段
第三阶段:自主行动阶段
设计模式:
关键问题: 用户如何可能逐步发展对此系统的信任?
设计持续的伙伴关系,而非孤立的交易。
从: 用户登录 → 完成任务 → 登出 → 系统遗忘
到: 系统持续保持对以下方面的感知:
实现模式:
关键问题: 用户真正想要实现的目标是什么?一个代理系统如何能更有效地帮助他们实现这些目标?
从: 显式设计每一条可能的用户路径
到: 设计目标对齐机制,系统在其中动态构建路径
代理系统:
设计目标:
关键问题: 系统能否帮助用户实现他们尚未完全阐明的目标?
传统的用户体验指标(会话时长、转化率、点击次数)对于代理体验来说并不适用。
应衡量:
关系质量
复合价值
情境准确性
民主对齐
关键问题: 我们如何知道关系是否在变得更好,而不仅仅是更频繁?
提出以下问题:
针对你的具体用例:
从每个类别中选择 2-3 个指标:
跟踪这些指标数周/数月,而不仅仅是会话。
情境时间线
情感状态指示器
动态建议面板
推理显示(透明阶段)
置信度计
撤销/纠正模式
目标仪表盘
规划画布
偏好演进地图
问题: 将偏好存储为键值对(主题:深色),而不是演进模式
修复: 设计动态模型,理解行为模式、时间情境和随时间推移的演进
问题: 系统从一开始就是完全透明或完全自主的
修复: 设计三阶段信任演进,具有渐进式自主权和用户控制的信任级别
问题: 为基于关系的系统衡量会话时长和转化率
修复: 跟踪数周/数月的关系质量、复合价值、情境准确性
问题: 系统记住一切,没有用户控制
修复: 构建遗忘控制、记忆可视化和清晰的数据保留策略
问题: 专注于像素级完美的界面,而没有关系架构
修复: 从关系模型开始,然后设计支持持续伙伴关系的屏幕
问题: 当系统出错时,用户永久失去所有信任
修复: 设计清晰的纠正路径、系统从错误中学习以及信任恢复协议
查看 EXAMPLES.md 了解:
查看 REFERENCE.md 了解:
查看 CHECKLIST.md 了解:
| 传统用户体验 | Agentic UX |
|---|---|
| 会话时长 | 数月的关系深度 |
| 转化率 | 信任分数和委托舒适度 |
| 点击率 | 复合价值(第 6 个月 vs 第 1 个月) |
| 孤立的屏幕 | 连续的关系情境 |
| 静态偏好 | 动态模式演进 |
| 一刀切 | 个体自适应界面 |
| 显式导航 | 目标对齐的路径构建 |
| 二元权限 | 渐进式信任演进 |
屏幕永远重要。但关系更重要。
每周安装数
0
代码库
GitHub 星标数
89
首次出现
1970年1月1日
The paradigm shift from screen-centric to relationship-centric design.
Traditional UX optimizes individual screens and isolated interactions. Agentic UX designs for ongoing relationships where systems learn, remember, and evolve alongside users across sessions, devices, and contexts.
Core principle: Every interaction builds on learned preferences and user history. Systems don't just respond—they develop understanding that compounds over time.
Announce at start: "I'm using the Relationship Design skill to create an agentic, memory-aware interface that builds long-term relationships with users."
Use this skill when:
When NOT to use:
Old model: Store static preferences (theme: dark, language: EN)
New model: Maintain dynamic, evolving relationship models
Design for:
Key question: What would this experience look like if it remembered everything and got better over time?
Design interfaces that earn autonomy through graduated trust:
Stage 1: Transparency Phase
Stage 2: Selective Disclosure Phase
Stage 3: Autonomous Action Phase
Design patterns:
Key question: How might users develop trust with this system gradually?
Design ongoing partnerships, not isolated transactions.
From: User logs in → completes task → logs out → system forgets
To: System maintains continuous awareness of:
Implementation patterns:
Key question: What goals are users really trying to achieve, and how could an agentic system help them get there more effectively?
From: Design every possible user path explicitly
To: Design goal-alignment mechanisms where system dynamically constructs paths
Agentic systems:
Design for:
Key question: Can the system help users achieve goals they haven't fully articulated yet?
Traditional UX metrics (session duration, conversion rates, clicks) miss the point for agentic experiences.
Measure instead:
Relationship Quality
Compounding Value
Context Accuracy
Democratic Alignment
Key question: How do we know if the relationship is getting better, not just more frequent?
Ask these questions:
For your specific use case:
Choose 2-3 metrics from each category:
Track these over weeks/months, not just sessions.
Contextual Timeline
Emotional State Indicators
Dynamic Suggestions Panel
Reasoning Display (Transparency Phase)
Confidence Meter
Undo/Correct Patterns
Goal Dashboard
Planning Canvas
Preference Evolution Map
Problem: Storing preferences as key-value pairs (theme: dark) instead of evolving patterns
Fix: Design dynamic models that understand behavioral patterns, temporal context, and evolution over time
Problem: System is either fully transparent or fully autonomous from day one
Fix: Design three-stage trust evolution with gradual autonomy and user-controlled trust levels
Problem: Measuring session duration and conversion rates for relationship-based systems
Fix: Track relationship quality, compounding value, context accuracy over weeks/months
Problem: System remembers everything with no user control
Fix: Build forgetting controls, memory visualization, and clear data retention policies
Problem: Focusing on pixel-perfect interfaces without relationship architecture
Fix: Start with relationship model, then design screens that support ongoing partnership
Problem: When system makes mistakes, users lose all trust permanently
Fix: Design clear correction paths, system learning from mistakes, and trust recovery protocols
SeeEXAMPLES.md for:
SeeREFERENCE.md for:
SeeCHECKLIST.md for:
| Traditional UX | Agentic UX |
|---|---|
| Session duration | Relationship depth over months |
| Conversion rates | Trust scores and delegation comfort |
| Click-through rates | Compounding value (Month 6 vs Month 1) |
| Isolated screens | Continuous relationship context |
| Static preferences | Dynamic pattern evolution |
| One-size-fits-all | Individually adaptive interfaces |
| Explicit navigation | Goal-aligned path construction |
| Binary permissions | Graduated trust evolution |
The screens will always matter. But the relationships matter more.
Weekly Installs
0
Repository
GitHub Stars
89
First Seen
Jan 1, 1970
AI Elements:基于shadcn/ui的AI原生应用组件库,快速构建对话界面
60,400 周安装