skills.rest
funsaized
funsaizedCOMMUNITY
Explore all skills by this author →
Website𝕏@funsaizedGitHub

agent-browser

Automates deterministic browser tasks for AI agents using ref-based page interactions and multi-session control.

1|1|Updated May 1, 2026
Search Tags:#snapshot#refs#data-extraction#browser-automation#agent-browser#web-interaction
One-click install
npx skills add https://github.com/funsaized/claude-and-codex-settings --skill agent-browser-funsaized
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/funsaized/claude-and-codex-settings/tree/main/plugins/agent-browser/skills/agent-browser
Command: npx skills add https://github.com/funsaized/claude-and-codex-settings --skill agent-browser-funsaized

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The agent-browser Skill provides a compact, ref-based interface to automate browser tasks for AI agents, reducing manual workload and enabling reliable web interactions.

Core Features & Use Cases

  • Fast, deterministic browser automation using compact accessibility-tree snapshots and @eN refs
  • Interactions include open/navigate, click, fill, type, wait, capture screenshots, and extract data across multiple tabs/sessions
  • Use cases include data extraction, form automation, login flows, and testing web apps across complex sites

Quick Start

Open a URL, snapshot the page to capture refs, and perform actions using those refs to automate common web tasks.

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 tasks for AI agents using accessibility-tree snapshots?▼

Browser automation for AI agents uses compact accessibility-tree snapshots and @eN refs to enable deterministic page interactions. You open a URL, snapshot the page to capture refs, and perform actions like click, fill, or type using those refs to automate web tasks reliably.

What is ref-driven browser automation and how does it work?▼

Ref-driven browser automation provides a compact, ref-based interface for AI agents to interact with web pages. It captures accessibility-tree snapshots and assigns @eN refs to elements, allowing deterministic actions like clicking, filling, and data extraction across multiple sessions and tabs.

Can I extract data and automate form submissions across multiple browser tabs?▼

Yes, browser automation supports data extraction, form automation, and login flows across multiple tabs and sessions. You use ref-driven interactions to navigate, fill forms, and extract data from complex websites deterministically.

Do I need specific tooling to use ref-based browser automation?▼

Yes, ref-based browser automation requires the agent-browser tooling to function. It can also leverage optional references, scripts, and assets to extend workflows and enable more complex web interactions across sites.

What's the best way to handle deterministic web interactions for AI agents?▼

Ref-driven browser automation is ideal for deterministic web interactions because it uses compact accessibility-tree snapshots and @eN refs. This approach ensures reliable, repeatable actions like navigation, clicking, typing, and screenshot capture across complex sites.

Related Skills

Explore More Software Engineering Skills→
ryanmivilleryanmiville

agent-browser

Automate browser navigation, form filling, and data scraping with Playwright.

Community
Advanced
shynlee04shynlee04

web-search

Search the web and extract content via Tavily and Exa using the inference.sh CLI.

Community
Intermediate
vm0-aivm0-ai
1.1k

agent-browser

Automate browser navigation, form filling, screenshots, and data extraction.

Official
Intermediate
hemoryhemory

agent-browser

Displays a list of available commands for the specified tool.

Community
Intermediate