MCP Protocol

The Model Context Protocol (MCP) lets AI agents use external tools — file systems, databases, browsers, and custom services. GWC supports MCP natively in both Studio and Cloud.

How it works

  1. Enable a connector — go to Connectors in your dashboard and toggle on the MCP tools you need.
  2. Agent discovers tools — when your agent runs, GWC exposes the enabled MCP tools automatically.
  3. Agent calls tools — the agent decides which tool to use, sends a request, and gets structured results back.
  4. All local — MCP tool calls execute on your machine (Studio) or your cloud instance. Data never passes through GWC servers.

Built-in connectors

MCP Filesystem

Read and write local files from your AI agents.

MCP Browser

Web browsing, scraping, and screenshots.

MCP SQLite

Query and modify local SQLite databases.

MCP Web Search

Search the web and return structured results.

Custom MCP Server

Connect any MCP-compatible tool server.

Custom MCP server

Add your own MCP-compatible tool server in GWC Studio settings:

// Example: custom MCP server config in GWC Studio
{
  "mcpServers": {
    "my-tool": {
      "command": "node",
      "args": ["./my-mcp-server.js"],
      "env": {
        "API_KEY": "${env:MY_API_KEY}"
      }
    }
  }
}

Using MCP tools via SDK

Specify which MCP tools an agent can access when creating a task:

// Use MCP tools in your agent via the SDK
const gwc = new GWC({ apiKey: 'your-key' });

await gwc.task.create({
  name: 'Analyze database',
  agent: 'data-analyst',
  input: {
    prompt: 'Summarize the users table',
    mcpTools: ['mcp-sqlite'],
  },
});