agent-browser

Automate browser tasks via CLI for opening sites, filling forms, and extracting data.

Updated Feb 17, 2026
One-click install
npx skills add https://github.com/sergical/repo-architect --skill agent-browser-sergical
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/sergical/repo-architect/tree/main/.agents/skills/agent-browser
Command: npx skills add https://github.com/sergical/repo-architect --skill agent-browser-sergical

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

The agent-browser Skill enables automated browser interactions for AI agents, replacing manual web navigation, form filling, and data extraction with repeatable, scriptable actions.

Core Features & Use Cases

  • Browser automation CLI for interactions such as opening websites, filling forms, clicking controls, taking screenshots, scraping data, and testing web apps.
  • Provides a compact, ref-based interaction model via snapshot refs to minimize token usage, enabling AI agents to operate efficiently.
  • Use cases include end-to-end testing, data extraction from websites, automated form completion, and monitoring pages for changes.

Quick Start

Open a target website with agent-browser, snapshot the page to collect refs, and perform actions using those refs.

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 tasks like filling forms and taking screenshots with an AI agent?

Automate web tasks by driving browser interactions through a CLI, enabling AI agents to open websites, fill forms, click controls, take screenshots, and extract data with repeatable actions.

How does browser automation minimize token usage when an AI agent interacts with a page?

Browser automation minimizes token usage by using a compact, ref-based interaction model via snapshot refs, allowing AI agents to perform web actions efficiently without processing full page HTML.

What's the best way to get started with web automation for end-to-end testing?

Start web automation for end-to-end testing by opening a target website with the CLI, snapshotting the page to collect refs, and performing actions like clicking or typing using those collected refs.

Can I use this browser automation CLI for scraping data and monitoring pages for changes?

Yes, the browser automation CLI supports scraping data from websites and monitoring pages for changes, replacing manual web navigation and data extraction with scriptable actions.

Do I need any external dependencies to run automation scripts for web interactions?

No external dependencies are required to run automation scripts for web interactions, as the skill operates independently to drive browser-based tasks through its CLI interface.

Are there limitations to using snapshot refs for automated form completion?

Snapshot refs for automated form completion are limited to interactions identifiable by the snapshot model, meaning highly dynamic or complex web elements may require additional snapshotting to interact accurately.