Proto PersonaSAFE

Product Management skills framework built on battle-tested methods for Claude Code, Cowork, Codex, and AI agents.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: proto-persona
argument-hint: "[target user or segment]"
description: Create a proto-persona from current research, market signals, and team knowledge. Use when you need a working customer profile before deeper validation.
intent: >-
  Create an initial, assumption-based persona profile that synthesizes available user research, market data, and stakeholder knowledge into a working hypothesis about your target user. Use this to align teams early in product development, guide initial design decisions, and identify gaps in understanding that require validation through research.
type: component
theme: discovery-research
best_for:
  - "Getting a working customer profile before you can afford real research"
  - "Making team assumptions about users explicit and challengeable"
  - "Setting up a persona you intend to validate, not defend"
scenarios:
  - "We have no research budget yet but need a working customer profile to start"
  - "Everyone on the team pictures a different user and it's causing arguments"
estimated_time: "15-25 min"
---


## Purpose
Create an initial, assumption-based persona profile that synthesizes available user research, market data, and stakeholder knowledge into a working hypothesis about your target user. Use this to align teams early in product development, guide initial design decisions, and identify gaps in understanding that require validation through research.

This is not a validated persona—it's a "proto" (prototype) persona that evolves as you learn more. Think of it as a structured placeholder that prevents design-by-committee while acknowledging you don't have all the answers yet.

## Input

**Works best with:** The target user or segment you need a working profile for.
**Also useful:** Whatever signal exists — support themes, sales anecdotes, analytics, prior research — plus the decision the persona will guide.

Anything supplied with the invocation itself — text after the skill name, a pasted context dump, or an appended `ARGUMENTS:` line — counts as answers already given. Use it and skip whatever it covers; don't re-ask.

**Arriving empty-handed? That works too.** The skill asks who you think the user is and what you already know, then structures it and flags the assumptions needing validation.

**Example invocation:** `Proto-persona for solo bookkeepers adopting our receipt-scanning app — signal: 30 support tickets and 4 sales call notes.`

## Key Concepts

### What is a Proto-Persona?
A proto-persona is a lightweight, hypothesis-driven persona created from:
- **Existing research:** User interviews, surveys, analytics (if available)
- **Market data:** Industry reports, competitor analysis, demographic trends
- **Stakeholder knowledge:** Sales, support, and team insights
- **Informed assumptions:** Best guesses that need validation

### Proto vs. Validated Persona
| Proto-Persona | Validated Persona |
|---------------|-------------------|
| Created in hours/days | Created over weeks/months |
| Based on assumptions + limited research | Based on extensive user research |
| Used to align teams early | Used to guide detailed design |
| Evolves rapidly | Stable over time |
| Good enough to start | High confidence |

### Why Use Proto-Personas?
- **Speed:** Align teams quickly without waiting for months of research
- **Focus:** Provides a shared reference point for "who we're building for"
- **Hypothesis framing:** Makes assumptions explicit, which can then be validated
- **Prevents generic design:** "Design for everyone" = design for no one

### Anti-Patterns (What This Is NOT)
- **Not validated research:** Don't treat it as fact—it's a hypothesis
- **Not a replacement for user research:** Use it to *guide* research, not avoid it
- **Not demographic data alone:** Age and location don't explain behavior
- **Not permanent:** Proto-personas should evolve as you learn

### When to Use This
- Early-stage product development (before extensive user research)
- Kicking off a new feature or pivot
- Aligning stakeholders on target users
- Identifying research gaps (who do we need to interview?)

### When NOT to Use This
- After you've done extensive user research (create a validated persona instead)
- For mature products with known user segments (you should already have validated personas)
- As a substitute for quantitative data (proto-personas inform research; research validates them)

---

## Application

Use `template.md` for the full fill-in structure.

### Step 1: Gather Available Context
Before creating a proto-persona, collect:
- **User research:** Interview notes, survey results, support tickets
- **Analytics:** Usage data, demographics, behavioral patterns
- **Market data:** Industry reports, competitor user bases
- **Stakeholder insights:** Sales/support/CS teams who interact with users
- **Product context:** What problem are you solving? (reference `skills/problem-statement/SKILL.md`)

**If missing context:** Don't fabricate—note gaps and plan research to fill them.

---

### Step 2: Define the Persona's Identity

#### Name
Give the persona an **alliterative, memorable name** (makes it easier to reference).

```markdown
### Name
- [Alliterative name, e.g., "Manager Mike," "Startup Sarah," "Enterprise Emma"]
```

**Quality checks:**
- **Memorable:** Can the team recall it easily?
- **Not generic:** Avoid "User 1" or "Persona A"

---

#### Bio & Demographics
Describe who this person is in the real world.

```markdown
### Bio & Demographics
- [Age range]
- [Geographic location]
- [Social status (married, single, family, etc.)]
- [Online presence (active on LinkedIn, avoids social media, etc.)]
- [Leisure activities]
- [Career status (job title, industry, seniority)]
```

**Quality checks:**
- **Behavioral, not just demographic:** Don't stop at "30-40 years old, lives in SF"—add "Works remotely, active in Slack communities, juggles 3 side projects"
- **Context-relevant:** Only include demographics that influence product decisions

**Example:**
- "35-45 yea

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.