Quickly Connect to MCP Service via CLINE Plugin (Using MCP SQLITE as an Example)

Install VSCODE

Download and install the vscode package from the official website https://code.visualstudio.com/ .

Install CLINE

Find CLINE in the vscode extension list and select Install.

Configure CLINE to use XXXCloud QWQ API

1. Click the small robot icon on the left side of VSCODE

2. First-time users will be prompted to configure the chatbot’s API. If you have used it before, you can configure it by clicking the CLINE settings gear.

3. After configuring, you can perform a dialogue test

Seeing a reply indicates successful configuration.

Install UV and SQLITE

Windows users run the following in cmd

winget install --id=astral-sh.uv -e
winget install sqlite.sqlite

Mac users use

brew install uv
brew install sqlite3

Configure CLINE to use mcp-server-sqlite

Click the small robot icon to open CLINE

Select the small server icon at the top of the page

In the MCP Servers, select the installed tab, and Configure MCP Servers will be displayed below. Click to open CLINE’s MCP configuration file.

Modify the CLine MCP configuration file.

Windows reference modification:

{
  "mcpServers": {
    "sqlite_server" :{
        "command": "cmd.exe",
      "args": [
      "/c",
      "uvx",
      "mcp-server-sqlite",
      "--db-path",
      "D:\\tmp\\test.db"
    ]
    }
  }
}

Mac reference modification:

{
  "mcpServers": {
    "sqlite_server": {
      "command": "uvx",
      "args": [
        "mcp-server-sqlite",
        "--db-path",
        "/tmp/test.db"
      ]
    }
  }
}

Where “D:\tmp\test.db” and “/tmp/test.db” should be modified to the desired location for storing the sqlite database. It can be an existing db file, and if not, it will be created automatically.

A prompt will appear after saving the file.

A sqlite_server tab will appear under the installed tab. A green status indicates that it has been enabled.

Interaction Example

Note ⚠️: The following prompts should be completed in the same conversation without clearing the context, or emphasize using sqlite when asking a second time.

List databases

Enter the following command in the dialogue box and send:

What databases can you open?

The model will return the command that needs to be run and ask if you want to run it. Choose run command or approve to execute the command.

A task may involve multiple runs, and each run needs human approval.

If you want the model to run fully automatically, you can configure the auto approve option, but it may pose some security risks.

After running, you will receive the following response:

Create table

I need to create a list of employees recording an 11-digit employee ID, employee name, employee rank, and employee hire date. Please help me create it.

The model requests to execute the table creation command, choose approve.

This creates a table for storing data.

Insert random sample data

Please generate 20 random employee samples so I can check if the table meets expectations.

The model requests to execute the insertion command, choose approve.

The model requests to execute the select command to check if insertion was successful, choose approve.

After executing the select command, you can see random data successfully inserted into SQLite, completing the task.

Clear data

Please help me delete the data and list, I need an empty database.

The model requests to execute the list deletion command, choose approve.

The model requests to execute the list view function, choose approve.

The model confirms that the data has been cleared, completing the task.