@brainerce/mcp-server
Provides AI coding assistants with domain knowledge about the Brainerce headless commerce platform, including SDK reference, type definitions, business flows, critical rules, and live store capabilities.
README
@brainerce/mcp-server
Framework-agnostic domain knowledge for building on Brainerce, a headless commerce platform. This MCP server answers questions an AI coding assistant needs while building or extending a Brainerce-powered storefront — SDK reference, type definitions, business-flow sequencing, critical rules, and a given store's live capabilities.
It does not generate framework boilerplate. Use Next.js, Remix, Vite + React, Vue, Svelte, Astro, SolidStart, or whatever the surrounding project already uses — this server tells the agent what to build correctly, not how to lay out files. For opinionated scaffolding, see the companion create-brainerce-store CLI.
Install
Claude Code (plugin):
claude plugin install brainerce-mcp@claude-plugins-community
Any MCP client (stdio, manual config):
{
"mcpServers": {
"brainerce": {
"command": "npx",
"args": ["-y", "@brainerce/mcp-server"]
}
}
}
Tools
| Tool | Purpose |
|---|---|
get-store-capabilities |
Live features configured for a specific store (payment providers, OAuth, discounts, shipping, i18n). Call first — it tells the agent which conditional features to build. |
get-required-features |
Functional coverage checklist — what users must be able to do, not which files to create. Mandatory features must exist even when currently disabled in the store (they auto-hide; owners enable them later). |
get-critical-rules |
Do's and don'ts that cause production incidents if violated — SDK usage, auth token handling, checkout invariants, i18n, type safety. |
get-business-flows |
Non-negotiable ordered SDK call sequences: checkout, auth, password reset, OAuth, order confirmation, cart persistence, inventory reservation. |
get-sdk-docs |
SDK documentation by topic — setup, API usage, best practices. |
get-type-definitions |
TypeScript types for Product, Cart, Checkout, Order, etc. |
get-code-example |
Short, focused snippets for tricky SDK calls (payment confirmation, reservation countdown, OAuth redirect, variant selection). |
get-store-info |
Live store metadata (display name, currency, language) for a sales channel. |
get-integration-guide |
Step-by-step integration walkthrough — core (products/cart/checkout/payment/orders), optional (accounts/OAuth/promotions), or rules (validation/error codes/edge cases). |
Resources
sdk-types— the full Brainerce SDK type surface as a text resource.
Prompts
create-store— framework-agnostic brief for building a new Brainerce-powered store; instructs the agent to call the knowledge tools above before writing code.add-feature— focused prompt for adding one feature to an existing Brainerce store.
Self-hosting
pnpm build
pnpm start:stdio # stdio transport, for local MCP clients
pnpm start:http # Streamable HTTP + legacy SSE, for remote clients (default port 3100)
The HTTP server is a public, unauthenticated documentation oracle by design — everything behind it is public SDK/docs data. It applies allow-listed CORS, per-IP rate limiting, request size caps, and SSE session limits; see src/bin/http.ts for details.
Privacy
This server serves static SDK documentation and, for the store-scoped tools (get-store-info, get-store-capabilities, get-required-features), reads public store configuration via Brainerce's public API. It does not collect or store conversation data. See brainerce.com/privacy for the full policy.
License
MIT — see LICENSE.
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。