persona-insight

Persona Insight

Produce professional user persona cards that capture who the real users are — their roles, goals, daily context, pain points, and emotional landscape. Each persona is a standalone reference document that can inform design, planning, and prioritization decisions.

Expert Discovery

This skill uses dynamic expert loading. On every run:

1. Primary role: Always load product-manager.md (leads persona development)
2. Scan project experts: Glob studio/agents/*.md — load all custom experts the team has created
3. Scan built-in experts: Glob ${CLAUDE_SKILL_DIR}/../../agents/*.md — load shipped experts (skip any already loaded from project)
4. Match by relevance: From the loaded experts, select those relevant to the current domain context. Match by comparing the expert's ## Your Domain section and title against the user's input topic. Include 1-3 most relevant domain experts.
5. Skip template: Do not load _domain-expert-template.md — it's for creating new experts, not for consultation.

The primary role produces the initial artifact. Domain experts review and correct it in the Expert Review step.

Inputs

Accept one of:

• A description of the user or user type via $ARGUMENTS (e.g., "全职妈妈，两个孩子，关注营养")
• A domain workspace path — read studio/changes/{name}/event-storm.md to extract personas mentioned there
• A conversation context — infer from what the user has been discussing

Workflow

1. Gather context — understand the domain and user segment
2. Build persona — construct a structured persona card
3. Map empathy — understand what the user thinks, feels, sees, does
4. Expert review — domain experts verify and correct the persona
5. Validate — present to user for feedback
6. Write output — save persona document

Step 1: Gather Context

Extract or ask for:

• Domain: What industry/product/service?
• User segment: Which type of user? (may be multiple)
• Data sources: Any existing user research, support tickets, interview notes?

If working within a studio workspace, scan event-storm.md for personas already mentioned.

Step 2: Build Persona Card

For each persona, produce:

┌─────────────────────────────────────────────────┐
│  {Persona Name}                                 │
│  "{One-sentence quote that captures their voice}"│
├─────────────────────────────────────────────────┤
│                                                 │
│  Role:        {job title or life role}          │
│  Age range:   {typical age}                     │
│  Tech level:  {高/中/低}                         │
│  Context:     {daily situation, constraints}     │
│                                                 │
│  Goals:                                         │
│  1. {Primary goal — what they're trying to do}  │
│  2. {Secondary goal}                            │
│  3. {Underlying motivation — why they care}     │
│                                                 │
│  Pain points:                                   │
│  1. {Biggest frustration} — severity: 高         │
│  2. {Second frustration} — severity: 中          │
│  3. {Third frustration} — severity: 低           │
│                                                 │
│  Current tools:                                 │
│  - {What they use today and why}                │
│  - {Where current tools fail}                   │
│                                                 │
│  Success looks like:                            │
│  "{What 'mission accomplished' means to them}"  │
│                                                 │
└─────────────────────────────────────────────────┘

Use the product manager agent perspective to challenge and refine: "Is this persona based on real behavior patterns, or is it an idealized version?"

Step 3: Map Empathy

For each persona, produce an empathy map:

            ┌─────────────┐
            │   Thinks     │
            │ {beliefs,    │
            │  concerns,   │
            │  priorities} │
┌───────────┼─────────────┼───────────┐
│   Sees    │             │   Hears   │
│ {what they│   PERSONA   │ {advice,  │
│  observe  │             │  feedback,│
│  in their │             │  noise}   │
│  world}   │             │           │
└───────────┼─────────────┼───────────┘
            │   Does      │
            │ {actions,   │
            │  behaviors, │
            │  workarounds│
            └─────────────┘

Pain: {core frustration that cuts across all quadrants}
Gain: {core desire that would resolve the pain}

Step 4: Expert Review

If domain experts were discovered in Expert Discovery, use the Agent tool to have each relevant expert review the persona card and empathy map.

Give each expert subagent:

• Their agent definition (.md file content)
• The draft persona card and empathy map
• The instruction: "Review this persona from your domain expertise. Flag anything that's incorrect, oversimplified, or missing. Pay attention to: domain-specific pain points we missed, real-world constraints that affect this persona, and terminology that practitioners would actually use."

Incorporate corrections into the persona. Common improvements from domain experts:

• More precise pain points grounded in domain reality
• Constraints outsiders wouldn't know about (regulations, physical limitations, cultural factors)
• Corrected terminology (what practitioners actually call things)

If no relevant domain experts were found, skip this step.

Step 5: Validate

Present the persona card and empathy map to the user:

• "Does this match real users you know?"
• "What's missing or inaccurate?"
• "Are there distinct sub-segments within this persona?"

If the user identifies sub-segments, create separate persona cards for each.

Step 6: Write Output

If working within a studio workspace, write to:

studio/changes/{domain-or-plugin}/personas/{persona-slug}.md

If standalone (no workspace), write to the current directory or let the user choose.

Each persona file contains the full persona card + empathy map in markdown format.

If multiple personas were created, also write an index:

studio/changes/{name}/personas/README.md

Listing all personas with one-line summaries.