velora-mcp-demo

velora-mcp-demo

A free public demo of Velora's agentic-commerce MCP toolkit, hosting ~49 real tool schemas with three honesty levels: pure validations, read-only sample data, and refused side-effect tools.

Category
访问服务器

README

velora-mcp-demo

Free, always-on public demo of the Velora agentic-commerce MCP toolkit — hosted on Vercel's free Hobby plan, $0/month, no card, no billing cycle.

Serves tools.somosvelora.com — the production Cloud Run backend that used to live there is paused on purpose (cost), so this demo replaces the old 404 with something real: a live MCP server, Streamable HTTP transport, real tool schemas, and an explicit honesty design instead of a dead link.

Endpoints

  • /api/mcp — full tool surface (~49 tools, mirrored from the real production tool set).
  • /api/mcp/public — read-only subset with zero side effects (matches the URL already printed in the latam-tools repo).

Three honesty levels (see lib/shared/types.ts and lib/shared/register.ts)

  1. Pure — validate_cuit runs the real Argentine AFIP/ARCA check-digit algorithm for real, against whatever CUIT/CUIL you send it. No fixture involved.
  2. Read — tools that only look something up (catalog, customers, caja balance, sales metrics, shipping quotes, connection status, the open_* MCP Apps widgets, …) return a fixed, fictional dataset ("Almacén Demo SRL"). Every response carries a [DEMO] Datos de muestra — el backend de producción está pausado. banner and _demo: true in its JSON body — never mistakable for a real business.
  3. Refuse — tools with a real side effect in production (emit_invoice, emit_nota, register_sale, create_tracked_payment_link, send_whatsapp_text, caja_ciclo_caja, connect_mercadopago, …) return isError: true with an honest explanation. No fabricated CAE, payment id, tracking number, or credential — ever. The tool's name, description, and input schema are still the real production ones; only the execution is disabled.

Where the tool names/schemas came from

Every tool name, description, and Zod input shape here is copied/trimmed from the currently-registered tools in the private velora-track3-repo production codebase (src/lib/mcp/*.ts), verified file-by-file — not invented. One known gap, found while building this: the developer-facing /developers catalog page in that repo lists ~60 tool names (CUIT/CUIL/CBU validators, format_ars, split_iva, to_afip_date, build_afip_qr, describe_invoice_type, consultar_comprobante, consultar_padron) that are not actually registered as MCP tools anywhere in the codebase — they exist only on the marketing page. This demo mirrors the ~49 tools that are actually real and registered, not the marketing count.

Stack

Next.js (App Router) + mcp-handler 2.x + @modelcontextprotocol/server 2.x + Zod 4. Deployed on Vercel Hobby.

Local dev

npm install
npm run dev
# http://localhost:3000/api/mcp

Status

Demo endpoint — sample data only. The real Velora product lives at somosvelora.com; this repo is a portfolio/demo piece, not the production system.

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选