agent-browser

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

1|Updated Jul 4, 2026
One-click install
npx skills add https://github.com/trungenglish/SHOPWISE --skill agent-browser-trungenglish
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/trungenglish/SHOPWISE/tree/main/.agents/skills/agent-browser
Command: npx skills add https://github.com/trungenglish/SHOPWISE --skill agent-browser-trungenglish

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires agent-browser, and includes references (resource) components.

What problem does it solve? AI agents and automation scripts need a programmatic way to interact with websites—navigating pages, filling forms, clicking buttons, and extracting data—without parsing raw HTML or writing brittle CSS selectors. ## Core Features & Use Cases - Snapshot-Based Interaction: Get compact accessibility-tree snapshots with element refs (@e1, @e2) to click, fill, and select elements without CSS selectors. - Authentication & Sessions: Reuse login state via auth vault, persistent profiles, session names, or state files, with optional encryption at rest. - Capture & Verification: Take screenshots (including annotated and full-page), save PDFs, record video, and diff page states to verify actions. - Use Case: Automate an end-to-end test of a web app's signup flow: open the page, snapshot to discover form fields, fill credentials, submit, wait for the dashboard redirect, and capture a screenshot as evidence. ## Quick Start Ask the agent to open a website, fill in the login form with your credentials, and take a screenshot of the resulting dashboard page.

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 actions with agent-browser?▼

Open a URL with agent-browser open, run snapshot -i to get element refs like @e1, then use click, fill, or select with those refs. Re-snapshot after any navigation or DOM change because refs are invalidated when the page changes.

How do I handle login and authentication in browser automation?▼

Use the auth vault to store encrypted credentials and run agent-browser auth login, or save session state with state save and reload it later. You can also import cookies from a running Chrome instance via --auto-connect state save.

Can agent-browser run parallel browser sessions?▼

Yes, use the --session flag to create isolated sessions with independent cookies, storage, and tabs. Each named session runs concurrently, and agent-browser session list shows all active sessions.

Does agent-browser support mobile or iOS testing?▼

Yes, it supports iOS Simulator via the -p ios flag with device selection, tap, and swipe gestures. This requires macOS with Xcode and Appium with the xcuitest driver installed.

Why do I get a ref not found error after clicking?▼

Refs are invalidated whenever the page changes, including navigation, form submissions, and dynamic content updates. Run agent-browser snapshot -i again after any such action to obtain fresh refs before interacting further.

How do I restrict which websites the browser can visit?▼

Set AGENT_BROWSER_ALLOWED_DOMAINS to a comma-separated allowlist of trusted domains, which also blocks sub-resource requests to non-allowed origins. You can additionally use an action policy file to gate destructive commands.