skills.rest
chisuhua
chisuhuaCOMMUNITY
Explore all skills by this author →
GitHub

agent-browser

Automate browser navigation, element interaction, and content extraction via Playwright sessions.

Updated Mar 4, 2026
Search Tags:#browser automation#session management#web scraping#form filling#playwright#screenshot#headless browser
One-click install
npx skills add https://github.com/chisuhua/home --skill agent-browser-chisuhua
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/chisuhua/home/tree/main/.agents/skills.disabled/agent-browser
Command: npx skills add https://github.com/chisuhua/home --skill agent-browser-chisuhua

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? AI agents cannot natively browse the web, click buttons, or fill forms, which blocks tasks like web scraping, automated testing, and authenticated data collection. This Skill provides a session-based browser controlled through simple CLI commands with compact element references. ## Core Features & Use Cases - Element Interaction via @e Refs: Snapshot pages to get compact refs, then click, fill, drag, upload, or press keys without parsing full DOM trees. - Session Persistence: Maintain cookies, login state, and history across multi-step workflows, including OAuth and 2FA login flows. - Capture and Debugging: Take viewport or full-page screenshots, record session video with an optional cursor indicator, and execute arbitrary JavaScript. - Use Case: Log into a web app with stored credentials, navigate to a dashboard, extract table data with JavaScript, and save a full-page screenshot as evidence. ## Quick Start Ask the agent to open a website, list the interactive elements on the page, fill in a form, and take a screenshot of the result.

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 with an AI agent?▼

Open a session with the open function and a target URL, then use the returned @e element refs to click, fill, or select elements via the interact function. Re-snapshot after any navigation because refs are invalidated when the page changes.

How to fill and submit a web form using Playwright automation?▼

Snapshot the page to identify input refs, use the fill action with each ref and text value, then click the submit button ref. Verify success by re-snapshotting and checking the URL or page elements for confirmation indicators.

Does this browser automation support proxy servers and authentication?▼

Yes, the open function accepts proxy_url, proxy_username, and proxy_password fields for HTTP, HTTPS, and SOCKS5 proxies. Authenticated login flows including OAuth redirects and TOTP-based 2FA are documented in the references.

Why do I get an Unknown ref error when clicking an element?▼

Element refs are invalidated whenever the page navigates or the DOM changes, so stale refs trigger this error. Run the snapshot function again to get fresh refs, or use the snapshot returned inside each interact response.

Can I record video of a browser automation session?▼

Yes, set record_video to true when opening the session, optionally with show_cursor for a visible cursor indicator. The video file in WebM format is returned when you call the close function.

Related Skills

Explore More Software Engineering Skills→
MantisWareMantisWare

import

Import workspace templates from local files, URLs, or GitHub repositories with collision handling and secret prompts.

Community
Advanced
EwallywEwallyw

regex-vs-llm-structured-text

Creates a hybrid regex and LLM workflow for extracting structured text data.

Community
Intermediate
se1111thse1111th

web-search

Search the web and fetch page content as markdown.

Community
Intermediate
richardnguyen0715richardnguyen0715

scrapling

Scrape static and JavaScript-rendered web content with stealth and Cloudflare bypass.

Community
Advanced