# Relationship Design

> Design AI-first interfaces that build ongoing relationships through memory, trust evolution, and collaborative planning. Use when designing agentic UX, memory-aware interfaces, or relationship-centric product experiences. NOT for general UI/UX tasks.

- **Type:** Skill
- **Install:** `agentstack add skill-ariadoss-superskills-relationship-design`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [ariadoss](https://agentstack.voostack.com/s/ariadoss)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [ariadoss](https://github.com/ariadoss)
- **Source:** https://github.com/ariadoss/superskills/tree/main/design-skills/relationship-design

## Install

```sh
agentstack add skill-ariadoss-superskills-relationship-design
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Agentic UX Design — Relationship-Centric Interfaces

Design framework for AI-first interfaces that build ongoing partnerships, not isolated session interactions.

## Core Shift

Traditional UX optimizes for sessions. Relationship-centric UX optimizes for **sustained partnerships**.

"Every interaction builds on learned preferences and user history. Systems don't just respond — they develop understanding that compounds over time."

---

## Five Pillars

### 1. Memory Revolution
Moving beyond static preferences to capture:
- Behavioral patterns across sessions
- Emotional context (frustration, delight, confusion patterns)
- Temporal evolution (how needs change over time)
- Implicit signals (what users skip, ignore, repeat)

Design for: "What would this system know about the user after 3 months?"

### 2. Trust as Design Material

Three-stage trust model:
1. **Transparency** — system shows all reasoning, asks permission
2. **Selective Disclosure** — system explains only when deviating from expectations
3. **Autonomous Action** — system acts independently with clear recovery paths

Design for: trust that can be inspected, corrected, and rebuilt.

### 3. Relationship-Centric Architecture
Maintain continuous awareness of:
- User goals (not just current task)
- Communication patterns (terse vs. exploratory)
- Historical preferences
- Outstanding commitments

Design for: context that doesn't reset between sessions.

### 4. Dynamic Path Planning
Systems that construct workflows aligned with individual objectives:
- "What is this user actually trying to accomplish?"
- Adaptive suggestions based on observed patterns
- Proactive surfacing of relevant information

### 5. New Success Metrics

Replace traditional metrics with:
- **Relationship quality**: does the system understand the user better over time?
- **Compounding value**: does each interaction make future interactions more valuable?
- **Context accuracy**: is stored understanding correct and current?
- **Democratic alignment**: does the system serve the user's actual goals, not engagement goals?

---

## Design Process

1. **Understand relationship context** — what ongoing relationship is this product building?
2. **Map trust evolution** — how does trust develop over days, weeks, months?
3. **Design memory architecture** — what should be remembered, forgotten, surfaced?
4. **Build collaborative patterns** — how does the system and user co-create over time?
5. **Define success across timeframes** — week 1, month 1, month 6 metrics

---

## Common Pitfalls

- Treating memory as static settings rather than dynamic behavioral understanding
- Using traditional engagement metrics (DAU, session length) for relationship-based systems
- Designing trust as binary (trusted/not) rather than a spectrum
- Building for the first interaction, not the 100th
- Storing everything users say rather than what's actually useful to remember

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [ariadoss](https://github.com/ariadoss)
- **Source:** [ariadoss/superskills](https://github.com/ariadoss/superskills)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-ariadoss-superskills-relationship-design
- Seller: https://agentstack.voostack.com/s/ariadoss
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.