Brand Registry
MCP server that exposes brand identity guidelines (visual look and voice) as markdown, enabling LLMs to produce on-brand content.
README
Brand Registry
A public brand-guidelines registry that does two things from one Next.js app:
- MCP server — an endpoint LLM clients (Claude, Cursor, …) connect to that returns each brand's look (visual identity) and voice (tone & writing) as markdown, so a model can produce on-brand output.
- Showcase pages — public web pages that render each brand's identity visually (palette, type scale, sample UI) as a live example.
It's built to hold many brands: each brand is a folder of prose markdown, and new brands are added by dropping in files (plus a small visual theme entry).
How it works
- Content lives in
brands-content/. Each subdirectory is one brand (the folder name is itsslug) and contains prose markdown:overview.md— positioning, personality, tagline (first#heading = display name)look.md— color, type, logo, layout, imageryvoice.md— principles, tone, vocabulary, examples
- Visual themes live in
src/lib/themes.ts— the colors, fonts, and radius used to render each brand's showcase. Content stays prose-only, so keep the theme values consistent withlook.mdby hand. - The MCP endpoint and the showcase pages both read the same content via
src/lib/brands.ts.
MCP surface
The endpoint is served at /mcp (stateless, read-only).
Tools
list_brands— every brand's slug, name, available aspects, and asset count. Call first.get_brand({ brand, aspect? })— a brand's markdown.aspectisoverview|look|voice|all(defaultall).get_brand_assets({ brand })— a brand's image assets as public URLs with descriptions, so a model can reference or embed them.
Resources
brand://{slug}/{aspect}— one markdown document per brand/aspect.brand://{slug}/assets— JSON manifest of the brand's image URLs (present only when the brand has assets). All enumerable viaresources/list.
Connect a client
{
"mcpServers": {
"brand-registry": {
"url": "https://<your-deployment>/mcp"
}
}
}
Locally the URL is http://localhost:3000/mcp.
Add a brand
- Create
brands-content/<slug>/withoverview.md,look.md,voice.md. (Any subset works; present files become the brand's aspects.) - Add a matching entry to
themesinsrc/lib/themes.tsso the showcase renders its real look. A brand with no theme entry falls back to a neutral default.
That's it — the home gallery, the /brands/<slug> page, and the MCP tools/resources
all pick it up automatically.
Add brand images
Images are served by URL (no base64). Per brand:
-
Drop image files in
brands-content/<slug>/assets/(SVG, PNG, JPG, WebP, AVIF, GIF). -
Describe them in
brands-content/<slug>/assets.jsonso clients know what each one is:{ "assets": [ { "file": "logo.svg", "name": "Primary logo", "kind": "logo", "description": "Full-color lockup; use on light backgrounds.", "background": "light" } ] }
Each file is served at /brands/<slug>/assets/<file> and surfaced through
get_brand_assets, the brand://<slug>/assets resource, and the showcase page.
Only files listed in assets.json are served (the manifest is an allowlist).
Absolute URLs — PUBLIC_BASE_URL
The MCP hands clients absolute asset URLs, so the server needs to know its public origin:
- On Vercel,
VERCEL_PROJECT_PRODUCTION_URLis used automatically. - Otherwise set
PUBLIC_BASE_URL(e.g.https://brands.example.com). - Local dev falls back to
http://localhost:3000.
Local development
npm install
npm run dev # http://localhost:3000
/— gallery of all brands + MCP connection info/brands/<slug>— a brand's full showcase/mcp— the MCP endpoint
Inspect the MCP endpoint interactively:
npx @modelcontextprotocol/inspector
# then connect to http://localhost:3000/mcp (Streamable HTTP)
Deploy (Vercel)
This is a standard Next.js App Router app, so Vercel needs no special config.
- Push this repo and import it in the Vercel dashboard (or run
vercel). - Deploy —
vercel --prod(or connect the repo for automatic deploys). - Your MCP endpoint is
https://<project>.vercel.app/mcp; the showcase is at the root.
next.config.ts bundles brands-content/ into the MCP function via
outputFileTracingIncludes so the markdown is readable at request time.
Tech
Next.js (App Router) · mcp-handler + @modelcontextprotocol/sdk · react-markdown.
The seed brands Nimbus and Verdant are fictional examples included to demonstrate the format.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.