three.ws Vision for VS Code

io.github.nirholas/vision-mcp

Image understanding for AI agents — analyze and describe any image via the three.ws pipeline.

client:VS Code transport:stdio runtime:npm

Install three.ws Vision in VS Code

.vscode/mcp.json

{
  "inputs": [
    {
      "type": "promptString",
      "id": "three-ws-base",
      "description": "THREE_WS_BASE",
      "password": true
    },
    {
      "type": "promptString",
      "id": "three-ws-timeout-ms",
      "description": "THREE_WS_TIMEOUT_MS",
      "password": true
    },
    {
      "type": "promptString",
      "id": "three-ws-api-key",
      "description": "THREE_WS_API_KEY",
      "password": true
    }
  ],
  "servers": {
    "vision-mcp": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@three-ws/vision-mcp"
      ],
      "env": {
        "THREE_WS_BASE": "${input:three-ws-base}",
        "THREE_WS_TIMEOUT_MS": "${input:three-ws-timeout-ms}",
        "THREE_WS_API_KEY": "${input:three-ws-api-key}"
      }
    }
  }
}

VS Code prompts for each input the first time the server starts, so no secret is written to the file. VS Code docs

three.ws Vision in other clients

three.ws Vision MCP server