import { Command } from "commander";
import { readFileSync } from "fs";
import { join } from "path";
import updateNotifier from "update-notifier";
export interface CliOptions {
port: number;
host: string;
ankiConnect: string;
ngrok: boolean;
}
function getPackageJson() {
try {
return JSON.parse(
readFileSync(join(__dirname, "../package.json"), "utf-8"),
);
} catch {
return { version: "0.0.0", name: "ankimcp" };
}
}
function getVersion(): string {
return getPackageJson().version;
}
export function checkForUpdates(): void {
updateNotifier({ pkg: getPackageJson() }).notify();
}
export function parseCliArgs(): CliOptions {
const program = new Command();
program
.name("ankimcp")
.description("AnkiMCP Server - Model Context Protocol server for Anki")
.version(getVersion())
.option(
"--stdio",
"Run in STDIO mode (for MCP clients like Cursor, Cline, Zed)",
)
.option("-p, --port <number>", "Port to listen on (HTTP mode)", "3000")
.option("-h, --host <address>", "Host to bind to (HTTP mode)", "127.0.0.1")
.option(
"-a, --anki-connect <url>",
"AnkiConnect URL",
"http://localhost:8765",
)
.option(
"--ngrok",
"Start ngrok tunnel (requires global ngrok installation)",
)
.addHelpText(
"after",
`
Transport Modes:
HTTP Mode (default): For web-based AI assistants (ChatGPT, Claude.ai)
STDIO Mode: For desktop MCP clients (Cursor, Cline, Zed)
Examples - HTTP Mode:
$ ankimcp # Use defaults
$ ankimcp --port 8080 # Custom port
$ ankimcp --host 0.0.0.0 --port 3000 # Listen on all interfaces
$ ankimcp --anki-connect http://localhost:8765
Examples - HTTP Mode with Ngrok:
$ ankimcp --ngrok # Start with ngrok tunnel
$ ankimcp --port 8080 --ngrok # Custom port + ngrok
$ ankimcp --host 0.0.0.0 --ngrok # Public host + ngrok
Examples - STDIO Mode:
$ ankimcp --stdio # For use with npx in MCP clients
# MCP client configuration (Cursor, Cline, Zed, etc.):
{
"mcpServers": {
"anki-mcp": {
"command": "npx",
"args": ["-y", "ankimcp", "--stdio"]
}
}
}
Ngrok Setup (one-time):
1. Install: npm install -g ngrok
2. Get auth token from: https://dashboard.ngrok.com/get-started/your-authtoken
3. Setup: ngrok config add-authtoken <your-token>
4. Run: ankimcp --ngrok
`,
);
program.parse();
const options = program.opts<CliOptions>();
return {
port: parseInt(options.port.toString(), 10),
host: options.host,
ankiConnect: options.ankiConnect,
ngrok: options.ngrok || false,
};
}
export function displayStartupBanner(
options: CliOptions,
ngrokUrl?: string,
): void {
const version = getVersion();
const title = `AnkiMCP HTTP Server v${version}`;
const padding = Math.floor((64 - title.length) / 2);
const paddedTitle =
" ".repeat(padding) + title + " ".repeat(64 - padding - title.length);
console.log(`
╔════════════════════════════════════════════════════════════════╗
║${paddedTitle}║
╚════════════════════════════════════════════════════════════════╝
🚀 Server running on: http://${options.host}:${options.port}
🔌 AnkiConnect URL: ${options.ankiConnect}${ngrokUrl ? `\n🌐 Ngrok tunnel: ${ngrokUrl}` : ""}
Configuration:
• Port: ${options.port} (override: --port 8080)
• Host: ${options.host} (override: --host 0.0.0.0)
• AnkiConnect: ${options.ankiConnect}
(override: --anki-connect http://localhost:8765)${ngrokUrl ? `\n • Ngrok tunnel: ${ngrokUrl}\n • Ngrok dashboard: http://localhost:4040` : ""}
${
!ngrokUrl
? `
Usage with ngrok:
1. Install: npm install -g ngrok
2. Setup: ngrok config add-authtoken <your-token>
3. Run: ankimcp --ngrok
`
: `
Share this URL with your AI assistant:
${ngrokUrl}
`
}
Run 'ankimcp --help' for more options.
`);
}
