brandguide/AI MCP
@bencium
Agents are taking over execution. When they write copy, generate visuals, or adapt messaging, they need structured brand knowledge — not PDFs and guesswork. brandguide/AI is the policy layer that makes agents brand-compatible by default.
Overview
What is brandguide/AI MCP?
brandguide/AI MCP is a Brand strategy knowledge base server for Claude Code, Cursor, Windsurf, and any MCP-compatible tool. It searches over 1,000 pages of brand methodology, design guidelines, and expert video transcripts directly from your development environment.
How to use brandguide/AI MCP?
Install and run it with npx brandguideai-mcp. No signup is needed. Configure it as an MCP server in your client (e.g., Claude Desktop, Cursor, Windsurf) by adding the command and arguments. Then use the search_brand_knowledge tool to ask brand strategy questions.
Key features of brandguide/AI MCP
- Searches a large brand strategy knowledge base.
- Single tool:
search_brand_knowledge. - Supports queries in English and Hungarian.
- Includes source citations from expert books and videos.
- Free tier: 4 queries with no account required.
- Uses the same RAG pipeline as the web app.
Use cases of brandguide/AI MCP
- Research brand identity definitions while coding.
- Get design guideline references for brand colors.
- Find brand naming best practices during a project.
- Access expert video transcript excerpts on brand positioning.
FAQ from brandguide/AI MCP
What questions can I ask brandguide/AI MCP?
You can ask anything about brand strategy, such as "What is brand identity?", "How do I choose brand colors?", or "What makes a good brand name?" It also works in Hungarian.
How many free queries do I get?
Every user gets 4 free queries with no account or API key required. After 8 queries, the tool returns upgrade links.
Do I need an API key or signup?
No — a temporary session is created automatically on first use. Existing brandguide/AI subscribers can use their quota automatically.
How does the server work technically?
It connects to the brandguide/AI knowledge base via a partner API. Queries are processed by a RAG pipeline using Voyage AI embeddings, Supabase vector search, and Claude for response generation.
Where are credentials stored?
Credentials are stored locally at ~/.brandguide/credentials.json.
