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

# Build with AI

Swig supports two complementary AI workflows:

1. **Give AI agents wallet capabilities** with the Swig MCP server
2. **Make the docs AI-friendly** by exposing the docs themselves as an MCP server

## Path 1: AI agents that can use Swig

Use this path when you want an agent to create wallets, manage authorities, configure RPC, or execute transactions.

* Start with [Getting Started with Swig MCP](/tutorials/mcp)
* Use [Swig MCP Server](/reference/mcp-server) for tool, transport, and deployment details
* Treat the MCP server as part of the Swig Protocol SDK surface in `swig-ts`

## Path 2: AI apps that need Swig docs as context

Use this path when you want your assistant, IDE, or internal tooling to ingest the Swig docs directly.

* Start with [Swig Docs MCP Support](/mcp-support)
* Use the generated MCP server to expose the docs to your AI tooling

## Good AI use cases for Swig

* agent-driven wallet creation and account setup
* chat-to-action experiences for users
* role-scoped agent permissions
* multi-agent workflows with limited authorities
* AI copilots that answer product questions using your docs

## Where to go next

<CardGroup cols={2}>
  <Card title="Wallet Agents" icon="terminal" href="/tutorials/mcp">
    Connect AI agents to Swig wallet operations
  </Card>

  <Card title="MCP Server Reference" icon="book" href="/reference/mcp-server">
    Review tools, permissions, and deployment details
  </Card>

  <Card title="Docs MCP" icon="book" href="/mcp-support">
    Expose the Swig docs themselves to LLM tooling
  </Card>

  <Card title="Products & SDKs" icon="globe" href="/products-and-sdks">
    Understand where AI fits across the product surface
  </Card>
</CardGroup>