# Brand persona — <Brand>

<!-- send_file {filename: "brand-persona.md", title: "Brand persona", docKind: "brand-guide", taskId: <tenant epic id>, content: <this file, filled>} -->

**Tenant:** `<slug>` · **Site:** <https://…> · **Date:** <YYYY-MM-DD> · **Approved by owner:** yes (Round 2)

## Who is speaking

- **Byline:** <Named human, role> → schema.org `Person`, or <Organisation> → `Organization`
- **Publisher:** <Organisation>
- **Stance:** <how the brand speaks about itself — e.g. "a small team building the tool it needed">

## Voice

| Adjectives (be) | Anti-adjectives (never) |
|---|---|
| <playful> | <corporate> |
| <knowledge-hungry> | <salesy> |
| <present in the conversation> | <preachy> |

Rules of thumb (from the owner's own writing, 3–5 posts fetched in Round 1):
- <sentence length, first/second person, humour, jargon level>
- <how technical terms are introduced>
- <what an opening paragraph does>

## Sample paragraph (approved)

> <the paragraph drafted from the owner's writing and approved in Round 2 — this is the calibration sample every draft is checked against>

## Claims policy

- **Competitors:** <name them / never name them / compare only on facts we can cite>
- **Claims we make:** <only what the product does today; cite docs or the owner's brief>
- **Claims we never make:** <e.g. no HIPAA / compliance claims; no medical, legal or financial advice; no invented numbers, quotes or anecdotes>
- **Sources:** every factual claim links a source in the draft notes.

## Locales

| Locale | Depth | Notes |
|---|---|---|
| `en` | master | all posts |
| `<de>` | full | translated + localised examples |
| `<fr>` | reduced | title, summary, key sections |

## Do / don't examples

- Do: <one line>
- Don't: <one line>