---
title: AI Integration
description: How to use AI to build Jaspr apps.
---

---

## Documentation for AI

Jaspr provides an AI-friendly documentation overview as a standard [llms.txt](https://jaspr.site/llms.txt) file. This is useful for adding the documentation as context to AI prompts when building Jaspr apps.

Provide the following url to your AI coding agent to give it access to the Jaspr docs:

```
https://jaspr.site/llms.txt
```

## Agent Skills

Jaspr ships with a set of **[Agent Skills](https://agentskills.io/)** — structured instruction files that teach AI coding agents how to work with Jaspr effectively. Each skill covers a specific topic:

| Skill | Description |
| --- | --- |
| **jaspr-fundamentals** | Core concepts of writing Jaspr components and using HTML components. |
| **jaspr-pre-rendering-and-hydration** | Techniques for SSR/SSG, async data fetching, hydration, and dual entrypoints. |
| **jaspr-styling** | Using type-safe CSS-in-Dart for styling components, themes, and CSS properties. |
| **jaspr-js-interop** | Integrating with JavaScript libraries and browser-native APIs. |
| **jaspr-convert-html** | Converting existing HTML into Jaspr code using the CLI. |

### Installing Skills

Use the `jaspr install-skills` command to copy the skills into your project's agent directory:

```bash
jaspr install-skills
```

Skills are versioned and will only be updated when a new version of Jaspr includes changes.

The command will auto-detect your agent by looking for an existing skills directory (e.g. `.agents/skills/`, `.cursor/skills/`).
You can also specify an agent explicitly:

```bash
jaspr install-skills --agent cursor
```

Supported agents include `cursor`, `antigravity`, `claude-code`, `copilot`, `windsurf`, `cline`, `codex`, `opencode`, `continue`, and `command-code`, as well as `general` which will use the standard `.agents/skills` directory.