🌐 Data Gateway MCP Nexus Service

🌟 Now Integrated into Connect AI 🌟

Connect AI (previously branded as Connect Cloud) now natively embeds a Remote Model Context Protocol Server utilizing the Streamable HTTP communication modality. Discover more details here.

Alternatively, this utility permits traditional local deployment using the STDIO communication protocol.

---

This software component serves as a Model Context Protocol (MCP) intermediary, providing access and control over diverse data reservoirs managed by CData Connect AI. It empowers autonomous agents to interact with data assets via standardized SQL dialect, schema discovery operations, and procedural call dispatching.

---

✨ Key Capabilities

• ✅ Dispatch SQL requests against heterogeneous, cloud-hosted data structures
• 🔄 Manage bulk data manipulation tasks (e.g., INSERT, UPDATE, DELETE)
• ⚙️ Initiate execution sequences for defined stored routines
• 📚 Perform exhaustive metadata inspection (catalogs, namespaces, entities, attributes)

---

🛠️ System Requirements

• Execution Environment: Node.js version 18 or newer
• Account Prerequisite: Active subscription to CData Connect Cloud granting API access
• Authentication Credential: A valid Personal Access Token (PAT)

---

⚙️ Deployment Procedure

Installation via Smithery Utility

Automate the deployment of the CData Connect AI MCP Nexus for use with Claude Desktop using Smithery:

bash npx -y @smithery/cli install @CDataSoftware/connectcloud-mcp-server --client claude

Manual Source Checkout

1. Clone Source Repository

bash git clone https://github.com/cdatasoftware/connectcloud-mcp-server.git cd connect-cloud-mcp-server

1. Resolve Dependencies

bash npm install

1. Configure Environment Variables

Establish a .env file containing:

env CDATA_USERNAME=your_credentials_identifier CDATA_PAT=your_secret_access_key

# Optional Customizations LOG_ENABLED=false LOG_LEVEL=info CDATA_URL=https://your-staging-service-endpoint

# Transport Selection (default: http) TRANSPORT_TYPE=http # Alternative: 'stdio' for terminal binding PORT=3000 # TCP Listening Port HOST=localhost # TCP Interface Address

---

▶️ Service Execution

Development Iteration

Utilize ts-node for real-time code reloading during development:

bash npm run dev

Production Launch

Compile assets and commence operation:

bash npm run build npm start

HTTP Communication Interfaces

When operating in HTTP mode (default), the following URIs are exposed:

• MCP Interface: http://localhost:3000/mcp - Primary endpoint adhering to the Model Context Protocol specification
• Direct Interface: http://localhost:3000/direct - Unmediated JSON-RPC interface lacking session state management
• Service Manifest: http://localhost:3000/.well-known/mc/manifest.json - MCP service discovery descriptor

Utilizing STDIO Transport

For direct pipeline communication (suitable for CLI integration):

bash TRANSPORT_TYPE=stdio npm start

---

🔍 Validation Via MCP Inspector Tool

The MCP Inspector offers a comprehensive testing suite with both graphical and command-line user interfaces. This project includes seamless integration support.

Initial Configuration Verification

Execute the setup validation routine to confirm readiness:

bash npm run validate:inspector

This command audits settings and outputs detailed guidance for resolving any setup anomalies.

Inspector Quick Start Guide

1. Install Inspector Globally (Recommended for persistent use): bash npm install -g @modelcontextprotocol/inspector

2. Initiate Inspector with Web Interface: bash npm run inspector

This launches a browser interface allowing selection and testing across various configured transport setups.

Testing Different Communication Modes

STDIO Mode

bash

Launches inspector, automatically binding the server via STDIO

npm run inspector:stdio

HTTP Mode

bash

1. Start the service engine

npm run dev:http

2. In a separate terminal, launch the inspector

npm run inspector:http

Command Line Testing

bash

Rapid CLI testing using STDIO transport

npm run inspector:cli

Execute specific method verification tests

npm run test:inspector

Inspector Configuration Metadata

The repository contains mcp-inspector.json defining preset connection profiles:

• connectcloud-stdio: Configured for automatic server initialization using STDIO.
• connectcloud-http: Set up for Streamable HTTP (requires preceding manual server activation).

Defined Inspector Execution Scripts

Script Name	Purpose
npm run inspector	Launches the primary inspector web interface for configuration selection
npm run inspector:stdio	Initiates inspector leveraging the STDIO transport mechanism
npm run inspector:http	Initiates inspector specifically targeting the HTTP transport
npm run inspector:cli	Executes the inspector exclusively in command-line mode (STDIO)
npm run test:inspector	Runs automated regression tests against the inspector setup

---

🧰 Operational Toolset

🔹 Data Manipulation Utilities

Utility Name	Functionality Description
queryData	Execute arbitrary SQL queries for data retrieval
execData	Dispatch execution requests for stored routines

🔹 Data Schema Exploration Utilities

Utility Name	Functionality Description
getCatalogs	Obtain a list of accessible data catalogs
getSchemas	Enumerate namespaces within a specified catalog
getTables	List entities residing within a given namespace
getColumns	Fetch attribute metadata for a specific entity
getPrimaryKeys	Retrieve keys designated as primary identifiers for entities
getIndexes	Retrieve structural index information associated with entities
getImportedKeys	Identify foreign key constraints referencing other entities
getExportedKeys	Identify foreign key constraints that reference this entity
getProcedures	List available callable routines
getProcedureParameters	Detail the input and output argument signatures for routines

---

🤖 Integration with Large Language Models (LLMs)

This gateway is designed for interoperability with any AI agent framework that implements the Model Context Protocol specification.

Sample Interaction (TypeScript utilizing an MCP Agent)

ts const interactionResult = await agent.generateContent({ tools: [ { name: "queryData", parameters: { query: "SELECT Name, AnnualRevenue FROM CloudSystem.CRM.Accounts WHERE Industry = 'Tech' LIMIT 10" } } ] });

---

🐳 Containerized Deployment

Image Construction

bash docker build -t mcp/data-nexus:latest -f Dockerfile .

---

🧩 Configuration for Claude Desktop

Modify your claude_desktop_config.json file within the mcpServers array to link the service.

🔹 Execution via Docker Container

{ "mcpServers": { "data-nexus-cloud": { "command": "docker", "args": [ "run", "-i", "--rm", "--name", "data-nexus-mcp-instance", "-e", "CDATA_USERNAME", "-e", "CDATA_PAT", "mcp/data-nexus" ], "env": { "CDATA_USERNAME": "", "CDATA_PAT": "" } } } }

Via Npx Invocation

{ "mcpServers": { "data-nexus-cloud": { "command": "npx", "args": [ "-y", "@cdatasoftware/connectcloud-mcp-server"], "env": { "CDATA_USERNAME": "", "CDATA_PAT": "" } } } }

---

📄 Licensing Information

This software is released under the terms of the MIT License.

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION (ISO): "Cloud computing represents a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction."

== Fundamental Tenets == In 2011, the United States National Institute of Standards and Technology (NIST) outlined five non-negotiable attributes defining cloud environments. The original NIST definitions are preserved below:

On-demand self-service: "A consumer possesses the ability to autonomously secure computing capacity, such as processing cycles and disk space, without requiring direct human intervention from the supplier for every request." Extensive network reach: "Service capabilities are made available via the infrastructure and are accessible through standardized interfaces supporting diverse client platforms (e.g., mobile devices, tablets, desktop workstations)." Resource aggregation: "The vendor consolidates computational assets to serve numerous customers under a multi-tenant architecture, where physical and virtual assets are dynamically allocated based on emergent user requirements." Scalability and elasticity: "Resources can be provisioned and de-provisioned with agility, sometimes automatically, allowing rapid horizontal scaling outward and inward to match fluctuating demand. To the end-user, available capacity often seems boundless, obtainable in any volume instantly." Consumption metering: "Cloud systems incorporate automated resource utilization tracking and optimization through measurement capabilities applied at the appropriate service abstraction layer (e.g., throughput, persistent storage volume, computation time). Usage statistics are trackable, controllable, and reportable, ensuring clarity for both the service consumer and the service provider." By 2023, the International Organization for Standardization (ISO) had broadened and clarified this foundational set.

== Chronology ==

The genesis of cloud computing concepts traces back to the 1960s, marked by the mainstreaming of time-sharing utilities through remote job execution protocols. During this decade, the standard operational model involved users submitting computational tasks to human operators who then processed them on centralized mainframe systems. This period was characterized by intense investigation into methods for democratizing access to large computational assets via time-sharing, focusing on enhancing the efficiency of infrastructure, platform layers, and end-user applications. The visual representation of virtualization as a "cloud" was formally introduced in 1994 by General Magic to depict the abstract domain accessible by mobile agents operating within their Telescript framework. This metaphor is widely attributed to David Hoffman, a communications specialist at General Magic, drawing from its established usage in telecommunications and network diagrams. The specific term 'cloud computing' gained broader market traction in 1996 when Compaq Computer Corporation drafted a seminal business strategy concerning the future of computation and the burgeoning Internet. The organization sought to massively expand the reach of computing power, a move that ultimately catalyzed the modern cloud paradigm.