pm-persona-codifierlisted
Install: claude install-skill spikelab/multiplai-cc-mktplace
# pm-persona-codifier
Turn raw archetype thinking into canonical persona docs. The output is structured so that every downstream PM doc (PRDs, strategy memos, roadmaps) can reference personas by ID without re-litigating who they are.
**This is codification, not invention.** The skill does not invent personas from thin air. It needs source material: interview transcripts, JTBD synthesis output, user notes, prior product docs, or a clear hypothesis from the user. If the source is thin, the skill says so honestly rather than fabricating detail.
## Arguments
| Arg | Description | Default |
|-----|-------------|---------|
| **source** | File path, folder, comma-separated list, or `--from-conversation` to use context already in the conversation | *(required)* |
| `--n` | Expected number of personas (the skill will confirm or push back) | auto-detect |
| `--depth` | `quick` (snapshot + jobs + quotes per persona) or `full` (all 9 sections) | `full` |
## Workflow
### Step 1 — Context gather (max 5 questions, all at once)
Before reading sources, confirm with the user (skip questions the user already answered):
1. **Who's the audience?** Internal team / leadership / sales enablement / all three. (Affects what to emphasize.)
2. **How many personas do you expect?** Push back if > 5. More than 5 is almost always under-clustered.
3. **Are these buyers, users, or both?** B2B SaaS usually has both, and they're often different people.
4. **What's already locked in?** Existing person