brand-knowledge-builder

Use when building, updating, or adding to a project’s brand knowledge base. Also use when onboarding a new client, processing brand documents, transcribing brand interviews, importing pre-structured knowledge folders, or filling in audience profiles, messaging frameworks, or competitive analysis.

Model	Source
sonnet	pack: brand

Full Reference

Brand Knowledge Builder

Builds and maintains a per-project brand knowledge base in .claude/knowledge/. Accepts four input types — pre-structured folders, documents, audio recordings, or Socratic interview — extracts brand information, and writes 10 structured markdown sections that the brand-strategist agent consumes. Also generates brand.json with machine-readable colors, fonts, logos, and products.

The 10 sections: brand-identity, audience-profiles, voice-and-tone, messaging-framework, competitive-landscape, content-strategy, business-model, visual-identity, brand-dictionary, product-catalog.

Quick Reference

Item	Value
KB location	.claude/knowledge/
brand.json	Project root
Input types	Pre-structured folder, documents (PDF/text), audio, no sources (interview)
Audio sub-skill	armadillo:deepgram-transcription (required)
Update policy	Additive only — never overwrite existing content
Scope options	Agency only, client only, or both

Reference Index

I want to…	File
Detect and handle input source types (folders, docs, audio)	reference/input-sources.md
Run the Socratic interview to fill knowledge gaps	reference/interview-questions.md
Write KB files and generate brand.json	reference/kb-files-and-brand-json.md

Usage: Read the reference file matching your current task from the index above. Each file is self-contained with code examples and inline gotchas.

---

┏━ 🧠 brand-knowledge-builder ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓ ┃ your friendly armadillo is here to serve you ┃ ┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛