Skip to content
Back to skills

Agent Browser

ASecurity

Use agent-browser CLI to interact with web pages — navigate, click, fill forms, take screenshots, inspect accessibility tree, and verify UI on a headless server.

  • 6 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 23, 2026
ai-agentsbashtestingapi

Works with

  • cli
  • api

Security analysis

A100/100

Scanned September 23, 2026

npx -y skills add vutran1710/claudebox --skill agent-browser --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Agent Browser?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Agent Browser
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/vutran1710-agent-browser/badge)](https://www.skillsdirectory.com/skills/vutran1710-agent-browser)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: agent-browser
description: Use agent-browser CLI to interact with web pages — navigate, click, fill forms, take screenshots, inspect accessibility tree, and verify UI on a headless server.
---

# Agent Browser

Agent-browser is a headless browser automation CLI installed on this server. Use it to navigate, interact with, and verify web applications without a desktop environment.

## Quick Reference

```bash
# Open a URL
agent-browser open "http://localhost:3000"

# See page structure (accessibility tree with element refs)
agent-browser snapshot

# Take a screenshot
agent-browser screenshot /tmp/page.png

# Click an element by ref
agent-browser click @e5

# Fill a form field
agent-browser fill @e3 "test@example.com"

# Type text (keystrokes)
agent-browser type @e3 "hello"

# Select dropdown option
agent-browser select @e4 "option-value"

# Press a key
agent-browser press Enter

# Hover over element
agent-browser hover @e2

# Scroll
agent-browser scroll down
agent-browser scroll @e5

# Get element text
agent-browser get text @e1

# Get page title/URL
agent-browser get title
agent-browser get url

# Set viewport size
agent-browser set viewport 1280 720

# Emulate mobile device
agent-browser set device "iPhone 15"

# Monitor network requests
agent-browser network --filter "api/"

# Wait for element/navigation
agent-browser wait @e5
agent-browser wait --url "**/dashboard"

# Compare screenshots (visual regression)
agent-browser diff before.png after.png

# Close browser
agent-browser close
```

## Workflow: Verify a Page

```bash
agent-browser open "http://localhost:3000"
agent-browser set viewport 1280 720
agent-browser snapshot                    # inspect structure
agent-browser screenshot /tmp/home.png    # visual check
```

## Workflow: Test a Form

```bash
agent-browser open "http://localhost:3000/checkout"
agent-browser snapshot
agent-browser fill @e3 "user@example.com"
agent-browser fill @e5 "2"
agent-browser click @e8                   # submit
agent-browser wait --url "**/success"
agent-browser screenshot /tmp/result.png
```

## Workflow: Debug API Calls

```bash
agent-browser open "http://localhost:3000/dashboard"
agent-browser network --filter "api/"     # watch XHR requests
agent-browser click @e5                   # trigger action
agent-browser network                     # inspect requests/responses
```

## Tips

- Always run `agent-browser snapshot` after navigation to get fresh element refs (`@e1`, `@e2`, etc.)
- Element refs change after page navigation or major DOM updates — re-snapshot
- Use `agent-browser screenshot` to visually verify; read the image with the Read tool to see it
- For mobile testing: `agent-browser set device "iPhone 15"` before navigating
- `agent-browser close` when done to free resources

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…