autolaunch

import { autolaunch } from "@tynd/core/client";

Start at system boot.

enable(opts?): Promise<{ enabled: boolean }>

interface AutolaunchOptions {
  name?: string;       // display name (registry key / .plist label / .desktop Name)
  args?: string[];     // extra CLI args appended when the OS relaunches
}
 
const { enabled } = await autolaunch.enable({ args: ["--minimized"] });

disable(opts?): Promise<{ enabled: boolean }>

const { enabled } = await autolaunch.disable();
// enabled is the new state (false on success)

opts takes the same shape as enable — usually you can pass nothing.

isEnabled(opts?): Promise<boolean>

const on = await autolaunch.isEnabled();

Platform-specific storage

• Windows — HKCU\Software\Microsoft\Windows\CurrentVersion\Run registry entry.
• macOS — ~/Library/LaunchAgents/<identifier>.plist.
• Linux — ~/.config/autostart/<name>.desktop.

All user-scoped — no admin / UAC prompt.

Path registered

The registered command is whatever std::env::current_exe() resolves to at the time of enable(). Typically:

• Installed app → path to the installed binary.
• Dev binary → target/release/tynd-full.exe or similar.

Re-register after an upgrade if the binary path changes.

Notes

• No per-user vs per-machine toggle. All three OSes use the user-scoped autostart store.
• enable() is idempotent — calling twice doesn’t duplicate the entry.
• Linux autostart obeys the XDG spec — the .desktop file needs correct Exec= / Name= / Type=Application. autolaunch writes these for you.

Example — toggle via settings

import { autolaunch } from "@tynd/core/client";
 
async function setAutolaunch(enabled: boolean) {
  if (enabled) await autolaunch.enable({ args: ["--minimized"] });
  else          await autolaunch.disable();
}
 
const current = await autolaunch.isEnabled();

Related

• singleInstance — argv forwarding.