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

# Key Concepts at a Glance

> A cheat sheet of what makes RunAgent different in 1 page

> TL;DR: RunAgent turns your Python AI agents into multi-language APIs with zero infrastructure headaches. Write once in Python, use everywhere.

## The Magic in 4 Bullets

<Cards num={2}>
  <Card title="Secure Isolation" icon="shield">
    **Sandboxed execution** - Each agent runs in its own secure, isolated environment, preventing resource conflicts and security issues.
  </Card>

  <Card title="Automatic Scaling" icon="chart-line">
    **Serverless architecture** - Automatically scales from zero to thousands of requests without manual configuration.
  </Card>

  <Card title="Cross-Language SDKs" icon="code">
    **Native APIs** - Call the same agent from Python, JavaScript, Go, Rust with language-idiomatic interfaces.
  </Card>

  <Card title="Real-Time Streaming" icon="stream">
    **Built-in streaming** - Token-by-token streaming for interactive applications and long-form generation.
  </Card>
</Cards>

***

## Traditional Approach vs RunAgent

<AccordionGroup>
  <Accordion title="Traditional: Build Everything Yourself">
    * Spin up FastAPI server
    * Build WebSocket handlers for streaming
    * Set up authentication and rate limiting
    * Deploy and manage infrastructure
    * Handle scaling and load balancing
    * Write SDKs for each language
    * Manage security and sandboxing
    * **Result**: Weeks of infrastructure work before you can use your agent
  </Accordion>

  <Accordion title="RunAgent: Deploy in One Command">
    * Write Python function
    * Run `runagent serve .`
    * Call from any language immediately
    * **Result**: Agent is live and accessible in seconds
  </Accordion>
</AccordionGroup>

***

## Core Concepts Cheat Sheet

| Concept         | What It Means                                          | Why It Matters                                           |
| --------------- | ------------------------------------------------------ | -------------------------------------------------------- |
| **Entrypoints** | Python functions exposed as API endpoints              | Your business logic becomes accessible from any language |
| **Tags**        | Nicknames for your functions (e.g., "chat", "analyze") | Clean, memorable API names instead of function names     |
| **Streaming**   | Real-time token delivery as they are generated         | Interactive UIs, progress indicators, better UX          |
| **Isolation**   | Secure, containerized execution environments           | Security, resource isolation, fast cold starts           |
| **SDKs**        | Language-native clients that feel like local functions | No need to learn REST APIs or WebSocket protocols        |
| **Config**      | `runagent.config.json` defines your agent's interface  | Single source of truth for what your agent can do        |

***

## The Developer Experience

<Steps>
  <Step title="Write Python Agent">
    ```python theme={null}
    def chat_agent(message: str) -> str:
        return llm.chat(message)
    ```
  </Step>

  <Step title="Deploy with One Command">
    ```bash theme={null}
    runagent serve .
    ```
  </Step>

  <Step title="Call from Any Language">
    ```python theme={null}
    # Python
    result = client.run(message="Hello!")
    ```

    ```javascript theme={null}
    // JavaScript
    const result = await client.run({message: "Hello!"});
    ```

    ```go theme={null}
    // Go
    result, _ := client.Run(ctx, map[string]interface{}{"message": "Hello!"})
    ```
  </Step>
</Steps>

***

## What Makes This Different

<CardGroup cols={2}>
  <Card title="Agent-First Design" icon="robot">
    Built specifically for AI agents, not generic microservices. Handles streaming, state, and agent patterns out of the box.
  </Card>

  <Card title="Zero Infrastructure" icon="cloud">
    No Docker, Kubernetes, or cloud configuration needed. Focus on your agent logic, not deployment.
  </Card>

  <Card title="Production Ready" icon="shield">
    Security, scaling, monitoring, and error handling built-in from day one.
  </Card>

  <Card title="Language Agnostic" icon="globe">
    Your Python agent becomes a first-class citizen in every supported language ecosystem.
  </Card>
</CardGroup>

***

## Ready to Dive Deeper?

<Cards num={3}>
  <Card title="Deploy Your First Agent" href="/tutorials/deploy-your-first-agent" arrow>
    See the magic in action with a 5-minute tutorial
  </Card>

  <Card title="Core Concepts Deep Dive" href="/explanation/core-concepts" arrow>
    Understand entrypoints, streaming, and configuration in detail
  </Card>

  <Card title="Architecture Overview" href="/explanation/architecture-overview" arrow>
    Learn how RunAgent works under the hood
  </Card>
</Cards>

<Card title="Still have a question?" icon="circle-question" href="/components/columns">
  * Join our [Discord Community](https://discord.gg/Q9P9AdHVHz)
  * Email us: [hi@run-agent.ai](mailto:hi@run-agent.ai)
  * Follow us on [X](https://x.com/run_agent)
  * New here? [Sign up](https://run-agent.ai/dashboard)
</Card>