メニュー
Automate browser interactions using agent-browser CLI. Use for taking screenshots, testing deployed sites, checking mobile views, form testing, or browser automation tasks.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
MCP-first testing workflow for Xcode apps (iOS, iPadOS, macOS). Use when the user asks to build, test, run, or verify an Xcode project — including UI checks, smoke tests, regression tests, or simulator launches. Prefers Xcode MCP tools via mcpbridge; falls back to xcodebuild/simctl CLI when MCP is unavailable.
Look up Apple Developer Documentation (Swift, SwiftUI, HealthKit, UIKit, etc.) and WWDC session transcripts using the sosumi CLI tool. Use when working with Swift/iOS code and need to check API signatures, find documentation, or understand Apple frameworks.
Generate user-friendly changelogs from git commits. Transforms technical commits into clear release notes.
Request and perform code reviews. Guidelines for reviewing code quality, finding bugs, and ensuring best practices.
Database management with Prisma ORM. Use for migrations, schema management, seeding, and database queries.
Docker and Docker Compose commands for containerization. Use for building images, running containers, managing volumes, and orchestrating services.
| name | agent-browser |
| description | Automate browser interactions using agent-browser CLI. Use for taking screenshots, testing deployed sites, checking mobile views, form testing, or browser automation tasks. |
| allowed-tools | Bash, Read |
| user-invocable | true |
Automate browser interactions for testing, screenshots, and web scraping using the agent-browser CLI tool from Vercel Labs.
npm install -g agent-browser
agent-browser install
The recommended pattern uses reference-based selection:
# 1. Navigate to page
agent-browser open "https://example.com"
# 2. Get interactive elements with refs
agent-browser snapshot -i
# 3. Interact using refs (@e1, @e2, etc.)
agent-browser click @e2
agent-browser fill @e3 "text"
agent-browser open <url> # Navigate to URL
agent-browser back # Go back
agent-browser forward # Go forward
agent-browser reload # Reload page
agent-browser snapshot # Full accessibility tree
agent-browser snapshot -i # Interactive elements only
agent-browser snapshot -c # Compact (remove empty elements)
agent-browser snapshot -d 3 # Limit depth to 3
agent-browser snapshot --json # JSON output for parsing
agent-browser click @e1 # Click by ref
agent-browser click "#submit" # Click by CSS selector
agent-browser dblclick @e1 # Double-click
agent-browser fill @e3 "text" # Clear and fill input
agent-browser type @e3 "text" # Type into element
agent-browser hover @e1 # Hover element
agent-browser focus @e1 # Focus element
agent-browser check @e1 # Check checkbox
agent-browser uncheck @e1 # Uncheck checkbox
agent-browser select @e1 "value" # Select dropdown option
agent-browser drag @e1 @e2 # Drag and drop
agent-browser upload @e1 file.pdf # Upload file
agent-browser press Enter # Press key
agent-browser press Control+a # Key combination
agent-browser scroll down 500 # Scroll down 500px
agent-browser scroll up 500 # Scroll up 500px
agent-browser scrollintoview @e1 # Scroll element into view
agent-browser screenshot /tmp/shot.png # Screenshot
agent-browser screenshot --full /tmp/full.png # Full page
agent-browser pdf /tmp/page.pdf # Save as PDF
agent-browser get text @e1 # Get element text
agent-browser get html @e1 # Get element HTML
agent-browser get value @e1 # Get input value
agent-browser get attr href @e1 # Get attribute
agent-browser get title # Get page title
agent-browser get url # Get current URL
agent-browser is visible @e1 # Check visibility
agent-browser is enabled @e1 # Check if enabled
agent-browser is checked @e1 # Check if checked
agent-browser find role button click --name "Submit"
agent-browser find text "Click me" click
agent-browser find label "Email" fill "user@test.com"
agent-browser find placeholder "Search..." type "query"
agent-browser wait @e1 # Wait for element
agent-browser wait 3000 # Wait 3 seconds
agent-browser set viewport 1920 1080 # Set viewport size
agent-browser set device "iPhone 14" # Emulate device
agent-browser set media dark # Dark mode
agent-browser set media reduced-motion # Reduced motion
agent-browser set offline on # Offline mode
agent-browser set headers '{"Auth": "Bearer token"}'
agent-browser open url --session mytest # Named session
agent-browser click @e1 --session mytest # Use same session
agent-browser close --session mytest # Close session
agent-browser console # View console logs
agent-browser errors # View page errors
agent-browser highlight @e1 # Highlight element
agent-browser --headed open url # Show browser window
agent-browser open "https://mysite.com" --session mobile
agent-browser set device "iPhone 14" --session mobile
sleep 2
agent-browser screenshot /tmp/mobile.png --session mobile
agent-browser close --session mobile
agent-browser open "https://mysite.com/login" --session test
agent-browser snapshot -i --session test
# Output: textbox "Email" [ref=e1], textbox "Password" [ref=e2], button "Login" [ref=e3]
agent-browser fill @e1 "user@example.com" --session test
agent-browser fill @e2 "password123" --session test
agent-browser click @e3 --session test
agent-browser screenshot /tmp/after-login.png --session test
agent-browser open "https://mysite.com" --session scroll
sleep 2
agent-browser scroll down 1000 --session scroll
sleep 1
agent-browser screenshot /tmp/scrolled.png --session scroll
agent-browser install)--headed to show)--json flag for machine-readable output