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

# Build with AI

> Use AI assistants and coding agents to learn and integrate with Porto faster.

<Callout icon="vault" color="#ac96ff">
  You're viewing **Porto** documentation — a distinct product from Anchorage Digital's core custody platform. [Go to Anchorage Digital docs →](/knowledge-base/build-with-ai/overview)
</Callout>

You can explore these docs and build your Porto integration with AI assistants and coding agents. Ask questions in natural language, pull our documentation straight into your AI tool, and start from ready-made prompts—while keeping the guardrails that self-custody requires.

## What you can do

<CardGroup cols={2}>
  <Card title="Connect your AI assistant" icon="plug" href="/knowledge-base/porto/build-with-ai/connect-ai-assistant">
    Add our docs to Claude, Cursor, VS Code, and more through the hosted MCP server.
  </Card>

  <Card title="LLM-ready docs" icon="file-code" href="/knowledge-base/porto/build-with-ai/llm-ready-docs">
    Copy any page as Markdown, open it in your AI tool, or pull the whole site.
  </Card>

  <Card title="Ready-made prompts" icon="wand-magic-sparkles" href="/knowledge-base/porto/build-with-ai/prompts">
    Start common integrations from copy-paste prompts.
  </Card>

  <Card title="AI security and guardrails" icon="shield-halved" href="/knowledge-base/porto/build-with-ai/security">
    Keep AI agents safe when they touch your self-custodied assets.
  </Card>
</CardGroup>

## New to AI-assisted development?

An **AI coding assistant**—such as Claude, Cursor, GitHub Copilot, or Windsurf—writes and edits code alongside you. The **Model Context Protocol (MCP)** is an open standard that connects these assistants to outside knowledge and tools, like a universal adapter. The assistant is the **client**; a **server** supplies the knowledge or actions. Anchorage Digital hosts an MCP server for these docs, so your assistant can search and cite them directly instead of guessing.

## Recommended path

<Steps>
  <Step title="Pick an AI tool">
    Use Claude Code, Cursor, VS Code with Copilot, Windsurf, or the Claude and ChatGPT apps.
  </Step>

  <Step title="Connect the Porto docs">
    Add our MCP server, or pull pages as Markdown, so your assistant answers from the current docs rather than guesswork. See [Connect your AI assistant](/knowledge-base/porto/build-with-ai/connect-ai-assistant).
  </Step>

  <Step title="Start from a prompt">
    Use a ready-made prompt for your task. See [Ready-made prompts](/knowledge-base/porto/build-with-ai/prompts).
  </Step>

  <Step title="Build with guardrails">
    Use scoped, read-only keys first, and keep a human in the loop before anything moves assets. See [AI security and guardrails](/knowledge-base/porto/build-with-ai/security).
  </Step>
</Steps>

<Note>
  Porto puts you in self-custody of real assets. Before you let an AI agent act on your account, read [AI security and guardrails](/knowledge-base/porto/build-with-ai/security).
</Note>

## Related

* [Set up API access](/knowledge-base/porto/developers/setting-up) — Create permission groups and keys, and sign requests.
* [API reference](/knowledge-base/porto/api-reference/introduction) — The full REST API, generated from the OpenAPI spec.


## Related topics

- [Build with AI](/knowledge-base/build-with-ai/overview.md)
- [Use these docs with AI](/knowledge-base/platform/users/using-ai.md)
- [Setting up](/knowledge-base/platform/developers/setting-up.md)
- [Sandbox testing guide](/knowledge-base/platform/developers/sandbox/sandbox-testing-guide.md)