Skip to main content
The Model Context Protocol (MCP) is an open standard that lets AI assistants interact with external tools and data sources. Avala’s MCP server exposes your datasets, projects, and exports as tools that any MCP-compatible client can call, so you can manage your annotation workflow through natural language without switching context. This means you can ask Claude to “list all my datasets,” have Cursor create an export, or use VS Code to check project status — all powered by the same Avala API you already use.

​
Prerequisites

  • Node.js 18 or later installed on your machine
  • An Avala API key (generate one in Mission Control under Settings > API Keys)

​
Setup

​
Claude Desktop

Open your Claude Desktop configuration file:
  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
Add the Avala MCP server: 
{
  "mcpServers": {
    "avala": {
      "command": "npx",
      "args": ["-y", "@avala/mcp-server"],
      "env": {
        "AVALA_API_KEY": "your-api-key"
      }
    }
  }
}
Restart Claude Desktop. You should see the Avala tools available in the tool picker.

​
Cursor

Create or edit .cursor/mcp.json in your project root (or global config directory): 
{
  "mcpServers": {
    "avala": {
      "command": "npx",
      "args": ["-y", "@avala/mcp-server"],
      "env": {
        "AVALA_API_KEY": "your-api-key"
      }
    }
  }
}
Restart Cursor or reload the window. The Avala tools will appear in Cursor’s agent mode.

​
VS Code

Create or edit .vscode/mcp.json in your workspace: 
{
  "mcpServers": {
    "avala": {
      "command": "npx",
      "args": ["-y", "@avala/mcp-server"],
      "env": {
        "AVALA_API_KEY": "your-api-key"
      }
    }
  }
}
Make sure you have the GitHub Copilot extension installed with MCP support enabled. Restart VS Code to pick up the new server.

​
ChatGPT

ChatGPT supports MCP through its plugins and actions interface:
  1. Open ChatGPT and navigate to Settings > Actions.
  2. Select Add Action and choose MCP Server.
  3. Enter the server command: 
    npx -y @avala/mcp-server
  4. Set the environment variable AVALA_API_KEY to your API key.
  5. Save and enable the action.
ChatGPT’s MCP support may vary depending on your plan and region. Refer to OpenAI’s documentation for the latest setup instructions.

​
Available MCP Tools

Once connected, the following tools are available to the AI assistant:
ToolDescription
list_datasetsList all datasets in your workspace with their IDs, names, and asset counts.
get_datasetGet detailed information about a specific dataset including data type and item count.
list_projectsList all annotation projects with their status and progress.
get_projectGet full project details including configuration and current status.
create_exportTrigger a new export for a dataset or project.
get_export_statusCheck whether an export is still processing, completed, or failed.
list_exportsList all exports with their creation dates and download URLs.
get_workspace_statsGet a summary of workspace usage including dataset count and project count.

​
Example Prompts

Here are some natural-language prompts you can use with any MCP-compatible client: 
List all my datasets

Show me the details of dataset "autonomous-driving-v3"

What's the status of the "Lidar QA" project?

Create an export for project 770a9600-g40d-63f6-c938-668877660000

How many exports do I have?

Show my workspace stats
The AI assistant will call the appropriate Avala MCP tool, process the response, and present the results in a readable format. You can chain requests in a single conversation — for example, list datasets, pick one, then create an export.
The Avala MCP server is open source. Visit the GitHub repository to view the source code or contribute.