BrandKity MCP Server
Enables AI agents to create, manage, and publish brand kits with colors, typography, logos, and white-label branding via the BrandKity platform.
README
@brandkity/mcp — BrandKity MCP Server
Model Context Protocol server for BrandKity — create and manage brand kits from any AI agent (Claude Desktop, Cursor, Windsurf, or any MCP-compatible client).
Current Version: 1.4.0 — Now with white-label branding support (Pro+)
Quick Start
1. Get an API Key
- Sign in at brandkity.com
- Go to Settings → API Keys
- Click Generate New Key and copy the key (
bk_live_...)
Requires a Pro or Agency plan for tool execution.
2. Configure Your AI Client
Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (Mac) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"brandkity": {
"command": "npx",
"args": ["-y", "@brandkity/mcp"],
"env": {
"BRANDKITY_API_KEY": "bk_live_your_key_here"
}
}
}
}
Cursor
Edit .cursor/mcp.json:
{
"mcpServers": {
"brandkity": {
"command": "npx",
"args": ["-y", "@brandkity/mcp"],
"env": {
"BRANDKITY_API_KEY": "bk_live_your_key_here"
}
}
}
}
Windsurf
Edit ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"brandkity": {
"command": "npx",
"args": ["-y", "@brandkity/mcp"],
"env": {
"BRANDKITY_API_KEY": "bk_live_your_key_here"
}
}
}
}
3. Use It
Once configured, ask your AI agent to create a brand kit:
"Create a brand kit for Acme Corp with the accent color #E55B00. Add a Colors block with the primary palette (Orange Flame #E55B00, Midnight #1A1A2E, Canvas #FAF9F7) and a Typography block with Inter for headings and DM Sans for body. Upload the logos from /Users/me/acme/logos/."
Available Tools (22)
| Tool | Description |
|---|---|
| Workspace | |
get_workspace |
Get workspace info (plan, kit count, storage) |
| Files | |
upload_file |
Upload any local file to workspace storage → returns a public URL |
list_files |
List workspace files with type filter and pagination |
| Kits | |
list_kits |
List all brand kits (filter by draft/published/all) |
create_kit |
Create a new kit → returns kit_id |
get_kit |
Get a kit with all blocks and content |
update_kit |
Update kit settings (name, color, template, logo_url, cover_image_url, white-label fields) |
publish_kit |
Publish a kit → returns public URL |
unpublish_kit |
Unpublish a kit (reverts to draft) |
| Blocks | |
list_blocks |
List all blocks in a kit with IDs and types |
ensure_block |
Idempotent — returns existing block_id or creates a new block (preferred over add_block) |
add_block |
Add a block unconditionally (use ensure_block instead to prevent duplicates) |
update_block |
Update block name/visibility |
delete_block |
Permanently delete a block and all its content |
| Content | |
add_colors |
Add color swatches to a Colors block |
add_typography |
Add font entries to a Typography block |
set_brand_story |
Set rich text content (brand story, tone of voice) |
set_block_note |
Set the editorial note displayed above any block |
| Upload | |
upload_asset |
Upload a local file into a block (logos, visuals, videos, etc.) with auto-retry |
upload_assets_batch |
Upload multiple local files into the same block; deduplicates by file path |
upload_kit_logo |
Upload and set the kit's header logo |
upload_cover_image |
Upload and set the kit's cover image |
White-Label Branding (Pro+ Feature)
Customize your portal with custom favicon, social share image, and SEO metadata:
// Upload custom assets
const faviconUrl = await client.uploadFile('favicon.ico', faviconBuffer);
const ogImageUrl = await client.uploadFile('og-image.png', ogImageBuffer);
// Apply white-label branding
await client.updateKit('kit-id', {
og_title: 'Acme Corp Brand Guidelines',
og_description: 'Official brand assets and standards',
custom_favicon_url: faviconUrl,
og_image_url: ogImageUrl,
});
Fields:
og_title(string, max 100 chars) — SEO title for social shareog_description(string, max 300 chars) — SEO descriptioncustom_favicon_url(string) — CDN URL to favicon (ICO/PNG/SVG)og_image_url(string) — CDN URL to social share image (1200×630 px recommended)
Plan Requirements:
- Free/Starter: White-label fields are read-only
- Pro/Agency: Full read-write access
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
BRANDKITY_API_KEY |
Yes | — | Personal Access Token (bk_live_...) |
BRANDKITY_API_URL |
No | https://brandkity.com |
API base URL (for local dev) |
Typical Workflow
1. get_workspace → verify connection, check plan and storage
2. list_kits → confirm kit doesn't already exist
3. create_kit → returns kit_id
4. ensure_block → idempotent: returns existing block_id or creates a new one (for each block type)
5. add_colors → populate the Colors block
6. add_typography → populate the Typography block
7. upload_file → upload font/logo/cover files to workspace storage
8. upload_asset → upload logos, images, videos into blocks
9. set_brand_story → write the brand story in a rich_text block
10. set_block_note → add usage guidance to any block
11. publish_kit → make the portal live
Reliability Notes (v1.4.0)
- No duplicate blocks —
ensure_blockis idempotent. Re-running a workflow never creates duplicate blocks. - Auto-retry on uploads —
upload_assetandupload_fileretry up to 3 times on network errors with exponential backoff. - Size-aware timeouts — Upload timeout scales with file size (60 s base + 20 s per 10 MB, max 10 min). Large files like 64 MB video assets are handled reliably.
- Batch deduplication —
upload_assets_batchsilently skips duplicatefile_pathentries so the same file is never uploaded twice in one batch. - Agent instructions — The server now provides operating rules to AI clients at connection time, reducing duplicate operations from AI agents automatically.
- White-label URL resolution — CDN URLs in white-label fields are automatically resolved to asset IDs server-side; agents don't need to manage asset IDs directly.
- Storage tracking — All file uploads are tracked per workspace for accurate quota enforcement.
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。