agent-browser

Automate browser tasks with a CLI for navigation, forms, screenshots, and data extraction.

64|2|Updated Feb 8, 2018
One-click install
npx skills add https://github.com/yuki-yano/dotfiles --skill agent-browser-yuki-yano
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/yuki-yano/dotfiles/tree/main/.agents/skills/agent-browser
Command: npx skills add https://github.com/yuki-yano/dotfiles --skill agent-browser-yuki-yano

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Automate repetitive browser interactions for AI-assisted tasks by providing a lightweight CLI to navigate pages, fill forms, capture data, and test web apps with reliable, human-readable element references.

Core Features & Use Cases

  • Snapshot-driven interactions using compact @e refs to minimize token usage and maximize reliability.
  • Multi-session isolation with state persistence for authenticated or separate contexts.
  • Automated content capture and testing workflows, including form automation, data extraction, and visual verification.
  • Real-world use case: coordinate a login flow, fill a form, and verify a dashboard URL across sessions.

Quick Start

Open a target URL with agent-browser open <url>, snapshot to generate element refs, and then interact with the page using those refs to complete your task.

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 for AI-assisted tasks?

Browser automation for AI-assisted tasks is handled through a lightweight CLI that navigates pages, fills forms, captures data, and tests web apps using reliable, human-readable element references. It provides compact snapshot-driven interactions to drive automated workflows.

What is the best way to fill forms and extract data from web pages via CLI?

Form filling and data extraction via CLI is achieved by opening a target URL, generating a page snapshot to create compact element references, and then interacting with those references to complete the desired data capture or form automation workflow deterministically.

Can I run parallel browser sessions with isolated state for authenticated browsing?

Yes, parallel browser sessions with isolated state are supported for authenticated or separate browsing contexts. The tool provides multi-session isolation with per-session state persistence, enabling concurrent automation workflows without cross-session interference.

Do I need a specific environment to run web testing and browser automation tasks?

Browser automation requires a modern shell, a compatible browser, and web-page readiness checks to ensure safe, deterministic execution. These prerequisites guarantee reliable interactions when navigating pages, taking screenshots, or extracting data.

How does snapshot-driven browser automation minimize token usage?

Snapshot-driven automation minimizes token usage by generating compact @e element references from page snapshots. These human-readable refs replace verbose DOM structures, allowing AI-assisted tasks to interact with web elements reliably while maximizing efficiency.

Why use element references instead of CSS selectors for web testing automation?

Element references provide a compact, human-readable alternative to CSS selectors for web testing automation. By using snapshot-generated @e refs, the CLI ensures reliable interactions that are less brittle and more efficient for AI-assisted browser workflows.