MCP Tools
MCP is an open protocol that standardizes how applications provide context to LLMs. Think of MCP like a USB-C port for AI applications. Just as USB-C provides a standardized way to connect your devices to various peripherals and accessories, MCP provides a standardized way to connect AI models to different data sources and tools.
Why MCP?โ
MCP helps you build agents and complex workflows on top of LLMs. LLMs frequently need to integrate with data and tools, and MCP provides:
- A growing list of pre-built integrations that your LLM can directly plug into
- The flexibility to switch between LLM providers and vendors
- Best practices for securing your data within your infrastructure
How to Create?โ
In the Digital Expert workspace, select the MCP tool type, and click the card to create a new MCP toolset.
As shown in the image below, fill in the necessary fields such as the name, then enter the MCP service configuration into the Schema field. Click "Load Tools" to attempt to load the tool list with the current configuration.

In the open-source code version, MCP tools running with the Command option will trigger a security prompt: The Command will run directly on the backend server, which may pose security risks. After upgrading to the Pro version, Command tools will run securely in a Sandbox container.
After loading the tool list for the MCP toolset, users can configure whether individual tools are enabled. You can also set "Disable Tools by Default" meaning tools that are not explicitly enabled will be turned off by default.
Examplesโ
Sequential Thinking: An MCP server implementation that provides a tool for dynamic and reflective problem-solving through a structured thinking process.
{
"mcpServers": {
"sequential-thinking": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-sequential-thinking"
]
}
}
}Playwright: A Model Context Protocol (MCP) server that provides browser automation capabilities using Playwright. This server enables LLMs to interact with web pages through structured accessibility snapshots, bypassing the need for screenshots or visually-tuned models.
{
"mcpServers": {
"playwright": {
"transport": "stdio",
"command": "npx",
"args": [
"-y",
"@playwright/mcp@latest",
"--headless"
]
}
}
}Blender: BlenderMCP connects Blender to Claude AI through the Model Context Protocol (MCP), allowing Claude to directly interact with and control Blender. This integration enables prompt-assisted 3D modeling, scene creation, and manipulation.
{
"mcpServers": {
"blender": {
"command": "uvx",
"args": [
"blender-mcp"
]
}
}
}More MCP.so
Transportโ
MCP (Model Context Protocol) has two main communication modes: Standard Input/Output (stdio) command mode and HTTP Server-Sent Events (SSE) mode.
Standard Input/Output Modeโ
This is the standard input/output communication mode, where the server runs as a local process and interacts via standard input (stdin) and standard output (stdout). It is suitable for development and testing scenarios and does not require a network connection.
HTTP SSE Modeโ
This is the remote communication mode, where the client connects to the serverโs HTTP endpoint, receives real-time events via SSE, and sends commands via HTTP POST. It is suitable for applications requiring remote access and real-time data streaming.
Mode Comparisonโ
The table below summarizes the differences between the two modes to help understand their applicable scenarios and characteristics:
Feature | Standard Input/Output Mode | HTTP SSE Mode |
---|---|---|
Communication | Local stdin/stdout | HTTP POST (commands) + SSE (events) |
Use Case | Local development, testing | Remote access, real-time data flow |
Network Needs | Local server network | Connects to other servers via network |
Real-Time | Synchronous, suitable for blocking ops | Asynchronous, supports real-time event push |
Security | Local operation, lower network risks | Requires network authentication and authorization |
Example Tools | Quarkus MCP stdio extension | Supergateway, Cloudflare remote deployment |