skills.rest
chhpt
chhptCOMMUNITY
Explore all skills by this author →
GitHub

agent-browser

Automates browser navigation, form filling, screenshots, and data extraction via CLI commands.

Updated Aug 23, 2026
Search Tags:#screenshots#cli#browser automation#web testing#web scraping#form filling#e2e testing
One-click install
npx skills add https://github.com/chhpt/skills --skill agent-browser-chhpt
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/chhpt/skills/tree/main/skills/agent-browser
Command: npx skills add https://github.com/chhpt/skills --skill agent-browser-chhpt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Manually testing web pages, filling repetitive forms, and scraping page content is slow and error-prone. This Skill lets an AI agent drive a real browser through simple CLI commands, turning multi-step web interactions into scripted, repeatable operations. ## Core Features & Use Cases - Page Interaction via Refs: Snapshot a page to get interactive elements with refs like @e1, then click, fill, hover, or select using those refs. - Screenshots & Data Extraction: Capture full-page screenshots and extract element text, input values, page titles, and URLs, with optional JSON output for parsing. - Sessions & Auth State: Run parallel browser sessions and save/load authentication state to skip repeated logins. - Use Case: Automate an end-to-end login test by opening the login page, filling credentials, submitting the form, waiting for the dashboard redirect, and saving the authenticated state for later runs. ## Quick Start Open example.com in the browser, take an interactive snapshot, and click the first button you find.

Frequently Asked Questions about agent-browser

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

FAQPage Schema
How do I automate form filling in a browser from the command line?▼

Open the page with agent-browser open, run snapshot -i to list interactive elements with refs, then use fill @ref "value" for each input and click @ref on the submit button. Wait for navigation with wait --load networkidle before verifying the result.

How do I take a full-page screenshot of a website programmatically?▼

Navigate to the URL with agent-browser open, then run agent-browser screenshot --full to capture the entire page. Add a file path argument to save it directly to disk instead of stdout.

Can I reuse login sessions across browser automation runs?▼

Yes. After logging in once, run agent-browser state save auth.json to persist cookies and storage. In later sessions, load it with agent-browser state load auth.json before navigating to authenticated pages.

How do I run multiple browser sessions in parallel?▼

Prefix commands with --session followed by a name, such as agent-browser --session test1 open site-a.com. Each named session maintains an independent browser instance, and session list shows all active sessions.

Why does my click fail after the page navigates?▼

Element refs become stale after navigation or major DOM changes. Re-run snapshot -i to get fresh refs before interacting again, or use semantic locators like find role button click --name "Submit" which resolve at execution time.

Related Skills

Explore More Software Engineering Skills→
civitas-cerebrumcivitas-cerebrum
5

journey-mapping

Discovers web application pages and user flows via Playwright to produce a prioritized journey map.

Official
Advanced
apconwapconw
2.2k

report-generation

Generate end-to-end HTML data analysis reports with charts and tables.

Community
Advanced
jesusalainmmjesusalainmm

schemas

Extract and validate JSON-LD schema data from web pages.

Community
Advanced
cheatcode-aicheatcode-ai
15

webapp-testing

Automate local web application testing with Playwright and headless Chromium.

Official
Intermediate