Headless hydration before opening a screen

async function prepareAndOpen(conversationId: string) {
  const runtimeName = `conversation-${conversationId}-runtime`;

  await ThreadedRuntime.prewarm(runtimeName);
  await ThreadedRuntime.runHeadlessTask('hydrateConversation', {
    runtimeName,
    payload: {
      conversationId,
      limit: 50,
    },
  });

  navigation.navigate('Conversation', { conversationId });
}

The flow:

1. Prewarm. The runtime starts if not already running.
2. Hydrate. A headless task fetches the first page of messages and writes them through the shared store.
3. Navigate. The user lands on the conversation; the screen renders against an already-populated store.

The screen has no spinner because the data is already there when it mounts.

When to use this

• List + detail flows. Prewarm + hydrate from the list before navigating.
• Onboarding flows. Hydrate the next step's data while the user reads the current step.
• Deep links. When opening a deep link, prewarm and hydrate before mounting the target screen.

When not to use this

• For lazy data the user might never see — prewarm but don't hydrate, or skip both.
• For data with a freshness window of seconds — hydrate inside the screen so it can show stale-then-fresh.