Demo-MCPServer
by JuanLadinoMoreno
Overview
An enterprise-ready MCP server for managing fast-food system data with secure OAuth 2.1 authentication via Scalekit, integrating AI tools for data analysis and reporting.
Installation
npm startEnvironment Variables
- PORT
- SK_ENV_URL
- SK_CLIENT_ID
- SK_CLIENT_SECRET
- MCP_SERVER_ID
- BASE_URL
- BASE_URL_API
- PROTECTED_RESOURCE_METADATA
Security Notes
The server implements robust OAuth 2.1 authentication with Scalekit, including token validation and granular permission checks per tool, and handles `nbf` claims. However, it uses `cors({ origin: ["*"], credentials: true })`, which is a critical security vulnerability, allowing any domain to make credentialed requests and potentially exposing the API to CSRF or other attacks. Error logging with `JSON.stringify(error.response.data)` could also expose sensitive backend error details if not carefully managed.
Similar Servers
SageMCP
A scalable platform for hosting MCP servers with multi-tenant support, OAuth integration, and connector plugins for various services, deployed on Kubernetes.
scalekit-mcp-server
This server enables AI agents to interact with Scalekit's identity platform through the Model Context Protocol (MCP) for natural language identity management.
converse
Orchestrates and exposes various AI tools (chat, multi-model consensus, job management) over the Model Context Protocol, enabling local, persistent, and potentially asynchronous AI interactions across multiple Large Language Model (LLM) providers.
brainloop-mcp-server-v2
Enables AI agents to interact with the Brainloop personalized learning platform by providing structured access to course management, content creation, and progress tracking, using an MCP (Model Context Protocol) interface.