/brand-setup

description: Configure brand voice, terminology, compliance guardrails, and style guide for content production
argument-hint: "<brand name> [--source=url|document|manual]"

Brand Setup

> If you see unfamiliar placeholders or need to check which tools are connected, see [CONNECTORS.md](../CONNECTORS.md).

Quick Start (5 minutes)

**Most users only need these 3 things to get started:**

1. **Brand name** — Your company or client name 2. **Voice/tone** — Pick one: authoritative, conversational, technical, witty, warm, educational 3. **Industry** — Pick one: pharma, bfsi, healthcare, legal, technology, b2b_saas, ecommerce, consumer_goods, real_estate, education

That's it. Run `/contentforge:cf-style-guide` and answer these 3 questions. ContentForge creates a working brand profile in under 5 minutes.

**You can always add more later:**

• Terminology and banned words → `/contentforge:cf-style-guide --update [brand]`
• Compliance guardrails → same command
• Audience personas, competitors, content pillars, visual identity → same command
• Tracking backend (Google Sheets / Airtable / Local) → `/contentforge:cf-switch-backend`

**Don't let setup slow you down.** A minimal brand profile works. You'll get better results as you add detail over time.

---

Full Setup (When You're Ready for More)

Create or update a brand voice profile that the ContentForge pipeline uses for every piece of content it produces. Import from existing style guide documents, URLs, or build interactively. Captures tone, formality, personality traits, approved and banned terminology, compliance guardrails, and industry-specific requirements.

Trigger

User runs `/contentforge:brand-setup` or asks to set up a brand, configure brand voice, import a style guide, or onboard a new client.

Inputs

Gather the following from the user. If not provided, ask before proceeding:

1. **Brand name** — the name for this brand profile (used in `--brand=` across all skills)

2. **Style guide source** — one of:

• **URL** — public URL to a style guide page (Notion, Google Docs published link, website, Confluence)
• **Document** — path to a .docx or .pdf style guide file
• **Manual** — interactive mode where you provide voice, terminology, and guardrails step by step

3. **Import scope** (optional, default: all):

• `voice` — extract only voice and tone characteristics
• `terminology` — extract only approved/banned terms
• `guardrails` — extract only compliance requirements
• `all` — extract everything

Setup Methods

Method 1: Import from Style Guide (Recommended)

If the user has an existing style guide document or URL:

1. **Fetch and parse** the style guide (WebFetch for URLs, document parser for .docx/.pdf) 2. **Extract voice characteristics:**

• Tone (authoritative, conversational, technical, witty)
• Formality level (1-5 scale)
• Personality traits (e.g., "bold but not aggressive", "technical but accessible")
• Writing style patterns (sentence length, paragraph structure, use of questions)

3. **Identify terminology:**

• Approved terms and preferred spellings
• Banned/prohibited words and phrases
• Industry jargon (keep, simplify, or avoid)
• Acronym handling rules

4. **Parse compliance requirements:**

• Required disclaimers by content type
• Prohibited claims (superlatives, health claims, financial promises)
• Regulatory framework (HIPAA, GDPR, financial services, etc.)
• Sensitivity guidelines

Method 2: Interactive Setup

Walk through 3 sections:

Voice & Tone (5 questions)

1. "Describe your brand voice in 3 words" (e.g., bold, witty, professional) 2. "How formal is your communication?" (1=very casual, 5=very formal) 3. "Who is your reader? What do they expect?" (maps to audience expectations) 4. "Name a brand whose writing style you admire" (reference point) 5. "Should the content use first person (we), second person (you), or third person?"

Terminology (3 questions)

6. "Any specific terms you always use?" (product names, branded terms, preferred spellings) 7. "Any words or phrases to avoid?" (competitor names, outdated terms, banned language) 8. "How should industry jargon be handled — keep it, explain it, or avoid it?"

Compliance (2 questions)

9. "What industry are you in? Any regulatory requirements?" (infer HIPAA, GDPR, etc.) 10. "Any mandatory disclaimers or legal language required in content?"

Method 3: Quick Start (Minimal Input)

For users who want to start producing content immediately: 1. Brand name 2. One-sentence description of what the brand does 3. Pick a tone: authoritative / conversational / technical / witty

The pipeline will use these defaults and refine the profile as more content is produced.

Profile Storage

Profiles are saved as structured JSON and used automatically by every pipeline phase:

• Phase 3 (Drafter) — applies voice and terminology
• Phase 5 (Proofreader) — enforces compliance and restrictions
• Phase 6 (SEO) — uses approved terminology in meta tags
• Phase 6.5 (Humanizer) — applies personality profile
• Phase 7 (Reviewer) — scores brand compliance

Tracking & Delivery Backend (Step G)

After setting up voice, terminology, compliance, and key files, **choose where ContentForge tracks quality scores and delivers output files**. Do NOT ask the user to edit JSON manually — handle it conversationally.

Step G.0 — PROBE ENVIRONMENT + EXISTING CONFIG FIRST (v3.12.7+, expanded v3.12.9)

**Before** offering the three-option menu, run TWO probes back to back to see what the user has and where they're running.

**First, detect the runtime environment:**

python scripts/plugin-metadata.py --section environment

If `environment == "cowork-sandbox"`, this changes the default backend logic significantly. In Cowork:

• The "Local" backend writes to the Linux sandbox, NOT the user's host -- files vanish at session end
• This means Drive (via the Anthropic platform integration or a Pipe