Overview

Learn how to build AI agents with AskUI’s core capabilities

This section contains practical how-to guides for building AI agents with AskUI. These guides focus on the core product capabilities: element selection and data extraction.

What You’ll Learn

Element Selection

• How to select UI elements using natural language
• Working with locators for precise element targeting
• Using relative positioning to find elements
• Handling multiple similar elements

Data Extraction (Get Questions)

• Extracting yes/no answers from the UI
• Getting text content from specific elements
• Extracting structured data from tables
• Working with sidebars, tabs, and cards
• Retrieving numeric values

Prerequisites

Before starting these guides, make sure you have:

• AskUI Python SDK installed (pip install askui)
• Basic understanding of Python
• Access to AskUI workspace and credentials

Quick Start

from askui import VisionAgent

# Initialize your agent
with VisionAgent() as agent:
    # Your first element interaction
    agent.click("Login button")
    
    # Your first data extraction
    is_loaded = agent.get("Is the page fully loaded?", response_schema=bool)

Guide Structure

Each guide in this section follows a practical approach:

1. What you’ll build - Clear objective
2. Code examples - Ready-to-use snippets
3. Common patterns - Reusable solutions
4. Troubleshooting - Solutions to common issues

Choose a guide from the navigation to get started building your AI agent.

API Reference

For detailed information about method parameters, return types, and advanced usage:

• click() - Click on UI elements
• get() - Extract data from UI
• act() - Autonomous goal-oriented actions
• type() - Type text input
• keyboard() - Send keyboard input
• Full Vision Agent API - Complete reference