agent-browser

Automate browser tasks including navigation, form filling, screenshots, and extraction via @refs.

Updated Jan 5, 2025
One-click install
npx skills add https://github.com/IdkwhatImD0ing/PortfolioV3 --skill agent-browser-idkwhatimd0ing
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/IdkwhatImD0ing/PortfolioV3/tree/main/.agents/skills/agent-browser
Command: npx skills add https://github.com/IdkwhatImD0ing/PortfolioV3 --skill agent-browser-idkwhatimd0ing

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

agent-browser removes the friction of manually navigating websites by giving AI agents a reliable way to open pages, inspect elements, and interact with web apps through compact references instead of brittle selectors.

Core Features & Use Cases

  • Interactive Web Automation: Open pages, take snapshots, click, fill, scroll, upload files, and extract content from live websites.
  • Session and Auth Workflows: Save and restore browser state, handle login flows, reuse sessions, and manage multiple isolated browser contexts.
  • Testing and Debugging: Capture screenshots, record video, inspect network activity, and validate page behavior during browser-based workflows.

Quick Start

Open the target page with agent-browser, capture an interactive snapshot, and use the returned refs to click, fill, or extract the elements you need.

Frequently Asked Questions about agent-browser

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate browser interactions without using brittle CSS selectors?

Browser automation can be handled by capturing compact snapshots with @refs, allowing AI agents to click, fill, and extract elements through reliable references instead of brittle selectors.

Can I save and restore authenticated sessions for web testing across multiple runs?

Authenticated sessions can be saved and restored during web testing, allowing you to handle login flows, reuse browser state, and manage multiple isolated contexts across runs.

What is the best way to extract data from interactive web pages using an AI agent?

Data extraction from interactive web pages is best achieved by opening the target page, capturing an interactive snapshot, and using the returned refs to target and extract the specific content needed.

How do I capture screenshots and record network activity during browser automation?

Screenshots and network activity can be captured during browser automation workflows to validate page behavior, inspect network requests, and debug interactive web applications.

Does agent-browser support managing multiple isolated browser contexts simultaneously?

agent-browser supports managing multiple isolated browser contexts simultaneously, enabling state persistence and separate session management for distinct automated web workflows.

Why use ref-based control instead of traditional selectors for web testing?

Ref-based control removes the friction of manually navigating websites by providing reliable references to inspect elements and interact with web apps, avoiding brittle selectors that break on UI updates.