skills.rest
wesbragagt
wesbragagtCOMMUNITY
Explore all skills by this author →
Website𝕏@wesbragagtGitHub

agent-browser

Automates browser interactions for UI testing, form filling, and screenshot capture via CLI.

1|Updated May 4, 2026
Search Tags:#screenshots#cli#browser automation#form filling#ui testing#web debugging#agent-browser
One-click install
npx skills add https://github.com/wesbragagt/nixos-config --skill agent-browser-wesbragagt
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/wesbragagt/nixos-config/tree/main/home/skills/browser
Command: npx skills add https://github.com/wesbragagt/nixos-config --skill agent-browser-wesbragagt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Manually clicking through web pages to verify UI changes is slow and repetitive. This Skill lets you drive a real browser from the command line to test forms, uploads, navigation, and page state without opening a browser yourself. ## Core Features & Use Cases - Browser Automation: Navigate pages, click elements, fill inputs, select dropdown options, upload files, and scroll using CSS selectors or accessibility-tree refs. - State Verification & Debugging: Take screenshots, capture accessibility snapshots, inspect console logs and page errors, and check element visibility or attributes. - Session & Network Control: Manage tabs, cookies, local storage, request routing, and isolated sessions for parallel test runs. - Use Case: While developing a PDF upload dialog, open the page, click "Import from PDF", select an instruction set, upload a test file, wait for the success toast, and screenshot the result to confirm the workflow end to end. ## Quick Start Ask the agent to open your local dev server page, fill in the login form, and take a screenshot to verify the page renders correctly.

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 testing from the command line?▼

Use the agent-browser CLI: run agent-browser open <url> to start a session, then issue commands like click, fill, select, and screenshot against CSS selectors or snapshot refs. Each command runs headlessly and returns results to the terminal.

How do I test a file upload flow in a web app?▼

Navigate to the page with agent-browser open, then run agent-browser upload "input[type='file']" with the file path. Follow with a wait command for the success indicator and a screenshot to confirm the upload completed.

Why doesn't agent-browser --headed show a window on Linux?▼

The --headed flag does not produce a visible window on Linux because the daemon launches Chrome with --no-startup-window. Instead, launch system Chromium directly with DISPLAY=:1 chromium <url> for manual login, then resume headless automation.

How do I find the right selector for a page element?▼

Run agent-browser snapshot -i to dump the interactive accessibility tree with @ref identifiers, or use agent-browser find with role, text, label, or testid locators. Prefer data-testid attributes over generic selectors for stability.

Can agent-browser handle pages that require login?▼

Yes, but headless sessions are independent, so open a visible Chromium window for the user to log in first, then pass cookies via agent-browser cookies set or the --headers option to authenticate the automated session.

Related Skills

Explore More Software Engineering Skills→
monobaselabsmonobaselabs

dev-app

Launch the account frontend development server with Bun on port 3002.

Official
Basic
agwacomagwacom

qa-only

Automate browser-based QA testing and export structured reports with evidence.

Official
Advanced
genlayerlabsgenlayerlabs
10

direct-tests

Execute in-memory unit tests for GenLayer intelligent contracts with mocking and VM cheatcodes.

Official
Intermediate
partme-aipartme-ai
610

cypress

Guide Cypress end-to-end and component testing with CI/CD configuration.

Official
Intermediate