dev-browser

Automate browser navigation, interaction, and extraction with persistent page state.

10|Updated Mar 22, 2026
One-click install
npx skills add https://github.com/Lee-SiHyeon/oh-my-copilot --skill dev-browser-lee-sihyeon
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dev-browser
Source: https://github.com/Lee-SiHyeon/oh-my-copilot/tree/main/.github/skills/dev-browser
Command: npx skills add https://github.com/Lee-SiHyeon/oh-my-copilot --skill dev-browser-lee-sihyeon

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

dev-browser eliminates the manual, error-prone effort of interacting with complex web pages by letting you automate navigation, form filling, and data extraction while keeping page state across runs.

Core Features & Use Cases

  • Persistent page sessions: Maintain page state on a dev-browser server so multi-step tasks can be completed incrementally.
  • Deterministic scripting: Run focused scripts to navigate, click, fill, wait, and extract results without rewriting everything for each step.
  • ARIA snapshot interaction: Discover and target elements reliably using accessibility snapshots, then click or interact via snapshot references.
  • Screenshots for verification: Capture visible and full-page screenshots to confirm what the user would see and to debug UI changes.
  • Use Case: Automate an end-to-end workflow like navigating to a site, logging in, submitting a form, and extracting specific confirmation data from the resulting page.

Quick Start

Start the dev-browser server first, then write and run a small script that connects to the server, navigates to your target URL, and performs the next single action before disconnecting.

Frequently Asked Questions about dev-browser

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

FAQPage Schema
How do I automate web scraping while keeping persistent page state across multiple script runs?

Persistent browser automation maintains page state on a dev-browser server so multi-step scraping tasks can be completed incrementally without rewriting scripts. You connect to a running server, perform actions like navigation or extraction, and disconnect to preserve context.

How do I target web page elements reliably during browser automation?

Browser automation targets web page elements reliably using ARIA snapshots. You discover and reference elements through accessibility snapshots, then execute interactions like clicks or form filling based on those snapshot references for deterministic scripting.

Can I verify website testing results visually without manually checking the browser?

Website testing results can be verified visually using screenshot capture. Browser automation supports taking visible and full-page screenshots to confirm what the user would see and debug UI changes without manual browser inspection.

Do I need a running server to automate form filling and website workflows?

Form filling and website workflows require a running dev-browser server. You must start the server first, then write and run scripts that connect to it, navigate to your target URL, and perform single actions before explicitly disconnecting to preserve state.

How do I complete an end-to-end workflow like logging in and submitting a form incrementally?

End-to-end workflows like logging in and submitting forms are completed incrementally by running focused scripts. Each script connects to the persistent page session, executes one action such as navigating or filling, and explicitly disconnects so the page state survives for the next run.

What are the limitations of using persistent sessions for website testing?

Persistent sessions for website testing require explicit disconnects to preserve page state, meaning scripts must be carefully structured into single actions. Complex multi-step workflows need multiple script runs, and a server must be actively running for connections.