agent-browser

Automate browser interactions and data extraction using compact element references.

Updated Oct 29, 2018
One-click install
npx skills add https://github.com/LuisUrrutia/.dotfiles --skill agent-browser-luisurrutia
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/LuisUrrutia/.dotfiles/tree/main/tools/opencode/config/.config/opencode/skills/agent-browser
Command: npx skills add https://github.com/LuisUrrutia/.dotfiles --skill agent-browser-luisurrutia

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Automates browser interactions and data extraction by providing compact element references, reducing AI prompt size and manual scripting.

Core Features & Use Cases

  • Snapshot-based element refs enable reliable, low-context interactions for automation tasks.
  • Supports common web actions: open/navigate, click, fill, type, screenshot, and data extraction across pages.
  • Use Case: automate login flows, form submissions, QA checks, and content capture across dashboards.

Quick Start

Open a URL with agent-browser, snapshot the page to generate refs, then interact using those refs to fill fields and submit forms.

Frequently Asked Questions about agent-browser

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

FAQPage Schema
How do I automate web testing without consuming too much AI prompt context?

Browser automation uses snapshot-based element references to minimize AI prompt size. By generating compact refs for page elements, the tool allows reliable interactions like clicking and filling fields without bloating the context window.

What is the best way to automate form filling and login flows across dashboards?

Automating form filling and login flows involves opening a URL, snapshotting the page to generate element refs, and interacting using those refs. This approach supports reliable UI QA checks and form submissions across diverse websites.

Can I use Bash commands to interact with web pages for data extraction?

Yes, a Bash-based interface exposes commands like open, snapshot, click, fill, and screenshot. This allows you to navigate pages and perform data extraction directly through command-line interactions.

How does snapshot-based element interaction work for UI QA checks?

Snapshot-based interaction captures the page state and assigns compact references to elements. You use these refs to perform actions like click or fill, ensuring reliable UI QA checks while keeping prompt sizes small.

Do I need any dependencies to run browser automation tasks with compact element refs?

No dependencies are required to run browser automation tasks with compact element refs. The tool operates independently, requiring only a SKILL.md frontmatter with a name and description, along with optional references.