‹ 首页

user-journey

@aleksander-dytko · 收录于 1 周前

Build a persona-grounded user journey map for a product flow

适合你,如果你需要梳理用户在产品中的完整行为路径和痛点。

/ 下载安装
user-journey.skill双击,或拖进 Claude 桌面版 / Cowork,即完成安装↓ .skill↓ .zip
用别的 agent?下载 .zip 解压,把文件夹放进它的技能目录
Claude Code~/.claude/skills/(项目级 .claude/skills/)
Codex CLI~/.codex/skills/
Cursor自动读取上面两处目录
其他工具见其文档的「skills」目录;两个下载是同一份文件,只是名字不同
/ 通过 npx 安装 校验哈希
npx oh-my-skill add aleksander-dytko/ai-pm-workspace/user-journey
/ 通过 bash 安装
curl -fsSL https://oh-my-skill.com/install.sh | bash -s -- aleksander-dytko/ai-pm-workspace/user-journey
/ 已经装过?验证本机副本,不用重装
npx oh-my-skill verify aleksander-dytko/ai-pm-workspace/user-journey
安装目标可用 --agent / --scope 或 --to 明确指定;省略时只会在唯一已存在的 agent 目录上自动选择,零命中或多命中会停止并提示。content_hash 缺失或不一致均拒装。
31GitHub stars
~1.1K上下文体积 · 单文件
镜像托管

怎么用

技能原文 SKILL.md作者撰写 · MIT · a6caa0d

User Journey

You help map a user journey for a specific task or flow. The output is a markdown journey artifact that names the persona, walks through stages, and highlights friction, emotion, and opportunities.

Input

The user provides via $ARGUMENTS:

  • A short description of the task or flow (e.g., "first-time API integration", "renewing a subscription", "handling a support escalation")
Workflow
1. Pick the persona

Read research/personas/ to see what personas exist.

If one persona clearly matches: confirm with the user - "I'll ground this journey in [[persona]]. OK?"

If multiple could match or none exists: ask the user which persona to use via AskUserQuestion, offering personas from research/personas/ plus an "Other - I'll describe it" option.

If no personas exist yet: prompt the user to run /personalize --deep or describe the persona in a short paragraph. Build the journey with an inline persona description at the top.

2. Define the task and its boundaries

Clarify in ONE AskUserQuestion round (skip any that are already answered):

  • Starting point: what triggers the user to begin this journey? (A specific event, not "they want to use the product".)
  • End state: what does "done" look like? (A specific measurable outcome.)
  • Context: time pressure, skill level, tools available.
  • Happy path or full path: do we map the ideal flow or include edge cases?
3. Map the journey

Produce a table-and-prose journey map with these columns:

| Stage | What the user does | What the user thinks | What the user feels | Friction / opportunity | |-------|--------------------|-----------------------|----------------------|-------------------------|

Keep stages at the right level of granularity - for most journeys, 5-8 stages is the sweet spot. Fewer misses detail; more turns into a task list.

For each stage:

  • What the user does: concrete action.
  • What the user thinks: internal voice, first-person.
  • What the user feels: emotion in one or two words.
  • Friction / opportunity: what's slow, confusing, or a moment of truth - and what could be improved.

Use specific, first-person voice ("I paste my API key and...") over generic ("The user enters credentials...").

4. Highlight the 2-3 biggest moments of truth

A "moment of truth" is a stage where the user's experience makes or breaks their trust in the product. Below the table, call out the 2-3 most critical:

### Moments of truth

1. **[Stage name]**: Why it's critical, what good looks like, what bad looks like.
2. ...
5. Propose opportunities and experiments

In the last section, list 3-5 concrete things to improve the journey. For each:

  • What the opportunity is.
  • What experiment would validate the fix.
  • Rough effort (small / medium / large).

These feed into /opportunity-solution-tree or a follow-up epic.

6. File location and naming
  • Folder: research/journeys/. Create if it doesn't exist.
  • Filename: YYYY-MM-DD - [Persona] - [Flow].md. Example: 2026-04-20 - Dana (Backend Dev) - First API Integration.md.
  • Frontmatter: tags: Journey, persona: [[persona file]], created: YYYY-MM-DD.
7. Link the journey
  • Daily journal: under ## Notes.
  • Related epic: if this journey was built for a specific epic, add a link in that epic's ## Breakdown > Scope section.
  • Persona file: append a reference to the journey at the bottom of the persona's Research Sources section.
8. Output summary
✅ Journey mapped: [[YYYY-MM-DD - Persona - Flow]]
✅ Linked in today's journal
[if epic] ✅ Linked in [[Epic]]

Persona: [[persona]]
Stages: [N]

🎯 Moments of truth:
1. [Stage name]
2. [Stage name]

💡 Top opportunity: [One-line summary]

Next step: run `/opportunity-solution-tree` to explore solutions for the top opportunity.
Notes
  • First-person voice beats third-person. Readers feel the journey more when it's "I" instead of "the user".
  • Emotion matters. A journey map without emotion is just a flowchart.
  • One persona per journey. If a journey serves multiple personas, the stages and friction are usually different - build a separate map per persona and compare.
  • Not a full feature spec. The journey highlights where the problems are; the epic describes what to build about them.
按 MIT 许可原样转载,未经改动 · 在 GitHub 查看 →

评论

登录即可评论;带「已验证安装」的,是发布者名下有本店的安装或持有记录。