> ## Documentation Index
> Fetch the complete documentation index at: https://docs.skybridge.tech/llms.txt
> Use this file to discover all available pages before exploring further.

# Core Concepts

> Understand the five key concepts that power Skybridge: write once run everywhere, data flow, LLM context sync, fast iteration, and type safety.

<Info>
  We assume that you are already familiar with ChatGPT Apps, MCP Apps, and MCP Servers. If you're not, read [Fundamentals](/fundamentals) first.
</Info>

Five concepts power Skybridge:

<CardGroup cols={2}>
  <Card title="Write Once, Run Everywhere" icon="shuffle" href="/concepts/write-once-run-everywhere">
    Write once, run anywhere - ChatGPT, Claude, Goose, VSCode, etc.
  </Card>

  <Card title="Data Flow" icon="arrows-rotate" href="/concepts/data-flow">
    How data moves between server, host, and view
  </Card>

  <Card title="LLM Context Sync" icon="brain" href="/concepts/llm-context-sync">
    Keep the model aware of view state via `data-llm`
  </Card>

  <Card title="Fast Iteration" icon="bolt" href="/concepts/fast-iteration">
    Local development with instant feedback
  </Card>

  <Card title="Type Safety" icon="shield-check" href="/concepts/type-safety">
    End-to-end TypeScript types from server to view
  </Card>
</CardGroup>

<Tip>
  **Start with Write Once, Run Everywhere** — it explains how Skybridge abstracts runtime differences so you can focus on your app.
</Tip>