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

# Build with AI

> Use AI coding assistants to build avnu integrations faster

Build Starknet applications with AI-assisted development. Install the avnu skill and Context7 to get code examples, SDK patterns, and up-to-date documentation directly in your AI assistant.

## Claude Skill

The [avnu skill](https://github.com/avnu-labs/avnu-skill) gives Claude Code deep knowledge of avnu SDK patterns and working code examples for:

<CardGroup cols={3}>
  <Card title="Token Swaps" icon="arrow-right-arrow-left">
    Best-price routing across Starknet DEXs
  </Card>

  <Card title="DCA Orders" icon="calendar-clock">
    Automated recurring token purchases
  </Card>

  <Card title="Staking" icon="coins">
    STRK and BTC variant staking
  </Card>

  <Card title="Gasfree Transactions" icon="gas-pump">
    Sponsor gas fees for your users
  </Card>

  <Card title="Market Data" icon="chart-line">
    Prices, volumes, TVL from Impulse API
  </Card>

  <Card title="Error Handling" icon="bug-slash">
    Common issues and debugging patterns
  </Card>
</CardGroup>

### Install

<Tabs>
  <Tab title="skills.sh">
    ```bash theme={null}
    npx skills add avnu-labs/avnu-skill
    ```
  </Tab>

  <Tab title="Context7">
    ```bash theme={null}
    npx ctx7 skills install avnu-labs/avnu-skill
    ```
  </Tab>
</Tabs>

Once installed, Claude Code automatically activates the skill when you discuss avnu, Starknet DeFi, or related topics.

## Context7

[Context7](https://context7.com/) injects up-to-date avnu documentation directly into your prompts. The docs and SDK references are fetched automatically — just add "use context7" to any prompt.

### Install

<Tabs>
  <Tab title="Claude Code">
    ```bash theme={null}
    claude mcp add context7 -- npx -y @upstash/context7-mcp@latest
    ```
  </Tab>

  <Tab title="Cursor">
    Add to `~/.cursor/mcp.json`:

    ```json theme={null}
    {
      "mcpServers": {
        "context7": {
          "command": "npx",
          "args": ["-y", "@upstash/context7-mcp@latest"]
        }
      }
    }
    ```
  </Tab>

  <Tab title="Claude Desktop">
    Add to `claude_desktop_config.json`:

    ```json theme={null}
    {
      "mcpServers": {
        "context7": {
          "command": "npx",
          "args": ["-y", "@upstash/context7-mcp@latest"]
        }
      }
    }
    ```

    **Config location:**

    * macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
    * Windows: `%APPDATA%\Claude\claude_desktop_config.json`
  </Tab>
</Tabs>

### Usage

```text theme={null}
Build an auto-compounding STRK staking integration. use context7
```

## Example Prompts

Try these prompts with the avnu skill and Context7 installed:

<CardGroup cols={2}>
  <Card title="Token Swap" icon="arrow-right-arrow-left">
    "Build a token swap from ETH to USDC with best price routing on Starknet"
  </Card>

  <Card title="Recurring Buy" icon="calendar-clock">
    "Create a weekly DCA order that buys 50 USDC worth of ETH every Monday"
  </Card>

  <Card title="Staking" icon="coins">
    "Build an auto-compounding STRK staking integration with reward claiming"
  </Card>

  <Card title="Gasfree" icon="gas-pump">
    "Set up sponsored gas transactions so my users don't pay any fees"
  </Card>

  <Card title="Market Data" icon="chart-line">
    "Fetch current prices, 24h volume, and TVL for the STRK/USDC pair"
  </Card>

  <Card title="Full Integration" icon="layer-group">
    "Build a DeFi dashboard with swaps, staking, and portfolio tracking using avnu"
  </Card>
</CardGroup>

***

## LLMs.txt

For custom tooling, RAG pipelines, or other AI workflows, we provide machine-readable documentation following the [llms.txt standard](https://llmstxt.org/):

* [llms.txt](https://docs.avnu.fi/llms.txt) — Structured index of all pages
* [llms-full.txt](https://docs.avnu.fi/llms-full.txt) — Complete documentation content

***

<Card title="Developer Support" icon="paper-plane" href="https://t.me/avnu_developers">
  Questions about AI integration? Join the avnu developers Telegram
</Card>