Model Context Protocol (MCP) Integration

Extend Thox.ai capabilities with MCP servers for enhanced AI-assisted development.

Back to Articles

The Model Context Protocol (MCP) is an open standard that enables AI models to interact with external tools and data sources. Thox.ai supports MCP, allowing you to extend your AI assistant with custom capabilities like file access, database queries, and API integrations.

What is MCP?

MCP (Model Context Protocol) is a standardized protocol that allows AI models to:

Access External Tools

Read/write files, execute commands, query databases

Maintain Context

Understand your project structure and codebase

Secure Integration

Permission-based access with user confirmation

Extensible Design

Add custom MCP servers for your specific needs

Built-in MCP Servers

Thox.ai includes several pre-configured MCP servers:

Filesystem

Read, write, and manage files in your project directories

Read filesWrite filesList directoriesSearch files

Git

Interact with Git repositories for version control operations

Clone reposCommit changesView diffsBranch management

Database

Query and manage SQL databases directly from the AI

Execute queriesView schemaInsert/Update dataTransactions

Terminal

Execute shell commands and scripts securely

Run commandsProcess outputEnvironment variablesWorking directory

Configuring MCP

MCP Configuration File

Edit the MCP configuration at ~/.thox/mcp.json:

json
{
  "mcpServers": {
    "filesystem": {
      "command": "mcp-server-filesystem",
      "args": ["--root", "/home/thox/projects"],
      "env": {}
    },
    "git": {
      "command": "mcp-server-git",
      "args": [],
      "env": {}
    },
    "sqlite": {
      "command": "mcp-server-sqlite",
      "args": ["--db-path", "/home/thox/data/app.db"],
      "env": {}
    }
  }
}

Via CLI

Manage MCP servers using the CLI:

bash
# List configured MCP servers
thox mcp list

# Enable/disable an MCP server
thox mcp enable filesystem
thox mcp disable sqlite

# Add a new MCP server
thox mcp add my-server --command "mcp-server-custom" --args "--config /path/to/config"

Using MCP in Practice

Once configured, MCP tools are automatically available to the AI. Here are some examples:

Ask the AI to read a file:

"Read the contents of src/index.ts and explain what it does"

Request code changes:

"Add error handling to the fetchData function in utils/api.ts"

Database queries:

"Show me the last 10 users who signed up from the database"

Git operations:

"Create a new branch called feature/auth and commit the current changes"

Creating Custom MCP Servers

You can create custom MCP servers to integrate with your specific tools and services:

typescript
import { Server } from "@modelcontextprotocol/sdk/server";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio";

const server = new Server({
  name: "my-custom-server",
  version: "1.0.0",
});

// Define a tool
server.setRequestHandler("tools/list", async () => ({
  tools: [{
    name: "my_tool",
    description: "Does something useful",
    inputSchema: {
      type: "object",
      properties: {
        input: { type: "string", description: "Input parameter" }
      },
      required: ["input"]
    }
  }]
}));

// Handle tool calls
server.setRequestHandler("tools/call", async (request) => {
  if (request.params.name === "my_tool") {
    const result = await doSomethingUseful(request.params.arguments.input);
    return { content: [{ type: "text", text: result }] };
  }
});

const transport = new StdioServerTransport();
await server.connect(transport);

Security Considerations

Confirmation prompts

Sensitive operations (file writes, commands) require user confirmation by default.

Sandboxed execution

MCP servers run in isolated environments with limited permissions.

Review third-party servers

Only install MCP servers from trusted sources. Review their code and permissions before use.

Related Articles

OpenAI-Compatible API

Use existing SDKs with Thox.ai

CLI Command Reference

All available CLI commands

Need help with MCP?

Check the MCP documentation or contact support for integration assistance.

MCP DocumentationContact Support