Remote Browser

Remote Browser Automation for Sandboxed Agents

When to Use

Use this skill as needed to perform the specified automation task.

This skill is for agents running on sandboxed remote machines (cloud VMs, CI, coding agents) that need to control a browser. Install browser-use and drive a cloud browser — no local Chrome needed.

Setup

Remote-only install (recommended for sandboxed agents)

curl -fsSL https://browser-use.com/cli/install.sh | bash -s -- --remote-only

This configures browser-use to only use cloud browsers:

• No Chromium download (~300MB saved)
• browser-use open <url> automatically uses remote mode (no --browser flag needed)
• If API key is available, you can also pass it during install:

  curl -fsSL https://browser-use.com/cli/install.sh | bash -s -- --remote-only --api-key bu_xxx
  

Manual install (alternative)

pip install "browser-use[cli]"

# Install cloudflared for tunneling:
# macOS:
brew install cloudflared

# Linux:
curl -L https://github.com/cloudflare/cloudflared/releases/latest/download/cloudflared-linux-amd64 -o ~/.local/bin/cloudflared && chmod +x ~/.local/bin/cloudflared

# Windows:
winget install Cloudflare.cloudflared

Then configure your API key:

export BROWSER_USE_API_KEY=bu_xxx   # Required for cloud browser

Verify installation:

browser-use doctor

Core Workflow

When installed with --remote-only, commands automatically use the cloud browser — no --browser flag needed:

# Step 1: Start session (automatically uses remote mode)
browser-use open https://example.com
# Returns: url, live_url (view the browser in real-time)

# Step 2+: All subsequent commands use the existing session
browser-use state                   # Get page elements with indices
browser-use click 5                 # Click element by index
browser-use type "Hello World"      # Type into focused element
browser-use input 3 "text"          # Click element, then type
browser-use screenshot              # Take screenshot (base64)
browser-use screenshot page.png     # Save screenshot to file

# Done: Close the session
browser-use close                   # Close browser and release resources

Understanding Installation Modes

When only one mode is installed, it becomes the default and no --browser flag is needed.

Exposing Local Dev Servers

If you're running a dev server on the remote machine and need the cloud browser to reach it:

# Start your dev server
python -m http.server 3000 &

# Expose it via Cloudflare tunnel
browser-use tunnel 3000
# → url: https://abc.trycloudflare.com

# Now the cloud browser can reach your local server
browser-use open https://abc.trycloudflare.com

Tunnel commands:

browser-use tunnel <port>           # Start tunnel (returns URL)
browser-use tunnel <port>           # Idempotent - returns existing URL
browser-use tunnel list             # Show active tunnels
browser-use tunnel stop <port>      # Stop tunnel
browser-use tunnel stop --all       # Stop all tunnels

Note: Tunnels are independent of browser sessions. They persist across browser-use close and can be managed separately.

Cloudflared is installed by install.sh --remote-only. If missing, install manually (see Setup section).

Commands

Navigation

browser-use open <url>              # Navigate to URL
browser-use back                    # Go back in history
browser-use scroll down             # Scroll down
browser-use scroll up               # Scroll up
browser-use scroll down --amount 1000  # Scroll by specific pixels (default: 500)

Page State

browser-use state                   # Get URL, title, and clickable elements
browser-use screenshot              # Take screenshot (base64)
browser-use screenshot path.png     # Save screenshot to file
browser-use screenshot --full p.png # Full page screenshot

Interactions (use indices from state)

browser-use click <index>           # Click element
browser-use type "text"             # Type into focused element
browser-use input <index> "text"    # Click element, then type
browser-use keys "Enter"            # Send keyboard keys
browser-use keys "Control+a"        # Key combination
browser-use select <index> "option" # Select dropdown option
browser-use hover <index>           # Hover over element
browser-use dblclick <index>        # Double-click
browser-use rightclick <index>      # Right-click

JavaScript & Data

browser-use eval "document.title"   # Execute JavaScript
browser-use extract "all prices"    # Extract data using LLM
browser-use get title               # Get page title
browser-use get html                # Get page HTML
browser-use get html --selector "h1"  # Scoped HTML
browser-use get text <index>        # Get element text
browser-use get value <index>       # Get input value
browser-use get attributes <index>  # Get element attributes
browser-use get bbox <index>        # Get bounding box (x, y, width, height)

Wait Conditions

browser-use wait selector "h1"                         # Wait for element
browser-use wait selector ".loading" --state hidden    # Wait for element to disappear
browser-use wait text "Success"                        # Wait for text
browser-use wait selector "#btn" --timeout 5000        # Custom timeout (ms)

Cookies

browser-use cookies get             # Get all cookies
browser-use cookies get --url <url> # Get cookies for specific URL
browser-use cookies set <name> <val>  # Set a cookie
browser-use cookies set name val --domain .example.com --secure  # With options
browser-use cookies set name val --same-site Strict  # SameSite: Strict, Lax, None
browser-use cookies set name val --expires 1735689600  # Expiration timestamp
browser-use cookies clear           # Clear all cookies
browser-use cookies clear --url <url>  # Clear cookies for specific URL
browser-use cookies export <file>   # Export to JSON
browser-use cookies import <file>   # Import from JSON

Tab Management

browser-use switch <tab>            # Switch tab by index
browser-use close-tab               # Close current tab
browser-use close-tab <tab>         # Close specific tab

Python Execution (Persistent Session)

browser-use python "x = 42"           # Set variable
browser-use python "print(x)"         # Access variable (prints: 42)
browser-use python "print(browser.url)"  # Access browser object
browser-use python --vars             # Show defined variables
browser-use python --reset            # Clear namespace
browser-use python --file script.py   # Run Python file

The Python session maintains state across commands. The browser object provides:

• browser.url - Current page URL
• browser.title - Page title
• browser.html - Get page HTML
• browser.goto(url) - Navigate
• browser.click(index) - Click element
• browser.type(text) - Type text
• browser.input(index, text) - Click element, then type
• browser.keys(keys) - Send keyboard keys
• browser.screenshot(path) - Take screenshot
• browser.scroll(direction, amount) - Scroll page
• browser.back() - Go back in history
• browser.wait(seconds) - Sleep/pause execution
• browser.extract(query) - Extract data using LLM

Agent Tasks

browser-use run "Fill the contact form with test data"   # AI agent
browser-use run "Extract all product prices" --max-steps 50

# Specify LLM model
browser-use run "task" --llm gpt-4o
browser-use run "task" --llm claude-sonnet-4-20250514
browser-use run "task" --llm gemini-2.0-flash

# Proxy configuration (default: us)
browser-use run "task" --proxy-country gb    # UK proxy
browser-use run "task" --proxy-country de    # Germany proxy

# Session reuse (run multiple tasks in same browser session)
browser-use run "task 1" --keep-alive
# Returns: session_id: abc-123
browser-use run "task 2" --session-id abc-123

# Execution modes
browser-use run "task" --no-wait     # Async, returns task_id immediately
browser-use run "task" --wait        # Wait for completion
browser-use run "task" --stream      # Stream status updates
browser-use run "task" --flash       # Fast execution mode

# Advanced options
browser-use run "task" --thinking    # Extended reasoning mode
browser-use run "task" --vision      # Enable vision (default)
browser-use run "task" --no-vision   # Disable vision

# Use cloud profile (preserves cookies across sessions)
browser-use run "task" --profile <cloud-profile-id>

# Task configuration
browser-use run "task" --start-url https://example.com  # Start from specific URL
browser-use run "task" --allowed-domain example.com     # Restrict navigation (repeatable)
browser-use run "task" --metadata key=value             # Task metadata (repeatable)
browser-use run "task" --secret API_KEY=xxx             # Task secrets (repeatable)
browser-use run "task" --skill-id skill-123             # Enable skills (repeatable)

# Structured output and evaluation
browser-use run "task" --structured-output '{"type":"object"}'  # JSON schema for output
browser-use run "task" --judge                          # Enable judge mode
browser-use run "task" --judge-ground-truth "answer"    # Expected answer for judge

Task Management

Manage cloud tasks:

browser-use task list                     # List recent tasks
browser-use task list --limit 20          # Show more tasks
browser-use task list --status running    # Filter by status
browser-use task list --status finished
browser-use task list --session <id>      # Filter by session ID
browser-use task list --json              # JSON output

browser-use task status <task-id>         # Get task status (latest step only)
browser-use task status <task-id> -c      # Compact: all steps with reasoning
browser-use task status <task-id> -v      # Verbose: full details with URLs + actions
browser-use task status <task-id> --last 5   # Show only last 5 steps
browser-use task status <task-id> --step 3   # Show specific step number
browser-use task status <task-id> --reverse  # Show steps newest first
browser-use task status <task-id> --json

browser-use task stop <task-id>           # Stop a running task

browser-use task logs <task-id>           # Get task execution logs

Cloud Session Management

Manage cloud browser sessions:

browser-use session list                  # List cloud sessions
browser-use session list --limit 20       # Show more sessions
browser-use session list --status active  # Filter by status
browser-use session list --json           # JSON output

browser-use session get <session-id>      # Get session details + live URL
browser-use session get <session-id> --json

browser-use session stop <session-id>     # Stop a session
browser-use session stop --all            # Stop all active sessions

# Create a new cloud session manually
browser-use session create                          # Create with defaults
browser-use session create --profile <id>           # With cloud profile
browser-use session create --proxy-country gb       # With geographic proxy
browser-use session create --start-url https://example.com  # Start at URL
browser-use session create --screen-size 1920x1080  # Custom screen size
browser-use session create --keep-alive             # Keep session alive
browser-use session create --persist-memory         # Persist memory between tasks

# Share session publicly (for collaboration/debugging)
browser-use session share <session-id>    # Create public share URL
browser-use session share <session-id> --delete  # Delete public share

Cloud Profile Management

Cloud profiles store browser state (cookies) persistently across sessions. Use profiles to maintain login sessions.

browser-use profile list                  # List cloud profiles
browser-use profile list --page 2 --page-size 50  # Pagination
browser-use profile get <id>              # Get profile details
browser-use profile create                # Create new profile
browser-use profile create --name "My Profile"  # Create with name
browser-use profile update <id> --name "New Name"  # Rename profile
browser-use profile delete <id>           # Delete profile

Using profiles:

# Run task with profile (preserves cookies)
browser-use run "Log into site" --profile <profile-id> --keep-alive

# Create session with profile
browser-use session create --profile <profile-id>

# Open URL with profile
browser-use open https://example.com --profile <profile-id>

Import cookies to cloud profile:

# Export cookies from current session
browser-use cookies export /tmp/cookies.json

# Import to cloud profile
browser-use cookies import /tmp/cookies.json --profile <profile-id>

Running Subagents

Cloud sessions and tasks provide a powerful model for running subagents - autonomous browser agents that execute tasks in parallel.

Key Concepts

• Session = Agent: Each cloud session is a browser agent with its own state (cookies, tabs, history)
• Task = Work: Tasks are jobs given to an agent. An agent can run multiple tasks sequentially
• Parallel agents: Run multiple sessions simultaneously for parallel work
• Session reuse: While a session is alive, you can assign it more tasks
• Session lifecycle: Once stopped, a session cannot be revived - start a new one

Basic Subagent Workflow

# 1. Start a subagent task (creates new session automatically)
browser-use run "Search for AI news and summarize top 3 articles" --no-wait
# Returns: task_id: task-abc, session_id: sess-123

# 2. Check task progress
browser-use task status task-abc
# Shows: Status: running, or finished with output

# 3. View execution logs
browser-use task logs task-abc

Running Parallel Subagents

Launch multiple agents to work simultaneously:

# Start 3 parallel research agents
browser-use run "Research competitor A pricing" --no-wait
# → task_id: task-1, session_id: sess-a

browser-use run "Research competitor B pricing" --no-wait
# → task_id: task-2, session_id: sess-b

browser-use run "Research competitor C pricing" --no-wait
# → task_id: task-3, session_id: sess-c

# Monitor all running tasks
browser-use task list --status running
# Shows all 3 tasks with their status

# Check individual task results as they complete
browser-use task status task-1
browser-use task status task-2
browser-use task status task-3

Reusing an Agent for Multiple Tasks

Keep a session alive to run sequential tasks in the same browser context:

# Start first task, keep session alive
browser-use run "Log into example.com" --keep-alive --no-wait
# → task_id: task-1, session_id: sess-123

# Wait for login to complete...
browser-use task status task-1
# → Status: finished

# Give the same agent another task (reuses login session)
browser-use run "Navigate to settings and export data" --session-id sess-123 --no-wait
# → task_id: task-2, session_id: sess-123 (same session!)

# Agent retains cookies, login state, etc. from previous task

Managing Active Agents

# List all active agents (sessions)
browser-use session list --status active
# Shows: sess-123 [active], sess-456 [active], ...

# Get details on a specific agent
browser-use session get sess-123
# Shows: status, started time, live URL for viewing

# Stop a specific agent
browser-use session stop sess-123

# Stop all agents at once
browser-use session stop --all

Stopping Tasks vs Sessions

# Stop a running task (session may continue if --keep-alive was used)
browser-use task stop task-abc

# Stop an entire agent/session (terminates all its tasks)
browser-use session stop sess-123

Custom Agent Configuration

# Default: US proxy, auto LLM selection
browser-use run "task" --no-wait

# Explicit configuration
browser-use run "task" \
  --llm gpt-4o \
  --proxy-country gb \
  --keep-alive \
  --no-wait

# With cloud profile (preserves cookies across sessions)
browser-use run "task" --profile <profile-id> --no-wait

Monitoring Subagents

Task status is designed for token efficiency. Default output is minimal - only expand when needed:

Recommended workflow:

# 1. Launch task
browser-use run "task" --no-wait
# → task_id: abc-123

# 2. Poll with default (token efficient) - only latest step
browser-use task status abc-123
# ✅ abc-123... [finished] $0.009 15s
#   ... 1 earlier steps
#   2. I found the information and extracted...

# 3. ONLY IF task failed or need context: use --compact
browser-use task status abc-123 -c

# 4. ONLY IF debugging specific actions: use --verbose
browser-use task status abc-123 -v

For long tasks (50+ steps):

browser-use task status <id> -c --last 5   # Last 5 steps only
browser-use task status <id> -c --reverse  # Newest first
browser-use task status <id> -v --step 10  # Inspect specific step

Live view: Watch an agent work in real-time:

browser-use session get <session-id>
# → Live URL: https://live.browser-use.com?wss=...

Detect stuck tasks: If cost/duration stops increasing, the task may be stuck:

browser-use task status <task-id>
# 🔄 abc-123... [started] $0.009 45s  ← if cost doesn't change, task is stuck

Logs: Only available after task completes:

browser-use task logs <task-id>  # Works after task finishes

Cleanup

Always clean up sessions after parallel work:

# Stop all active agents
browser-use session stop --all

# Or stop specific sessions
browser-use session stop <session-id>

Troubleshooting

Session reuse fails after task stop: If you stop a task and try to reuse its session, the new task may get stuck at "created" status. Solution: create a new agent instead.

# This may fail:
browser-use task stop <task-id>
browser-use run "new task" --session-id <same-session>  # Might get stuck

# Do this instead:
browser-use run "new task" --profile <profile-id>  # Fresh session

Task stuck at "started":

• Check cost with task status - if not increasing, task is stuck
• View live URL with session get to see what's happening
• Stop the task and create a new agent

Sessions persist after tasks complete: Tasks finishing doesn't auto-stop sessions. Clean up manually:

browser-use session list --status active  # See lingering sessions
browser-use session stop --all            # Clean up

Session Management

browser-use sessions                # List active sessions
browser-use close                   # Close current session
browser-use close --all             # Close all sessions

Global Options

Common Patterns

Test a Local Dev Server with Cloud Browser

# Start dev server
npm run dev &  # localhost:3000

# Tunnel it
browser-use tunnel 3000
# → url: https://abc.trycloudflare.com

# Browse with cloud browser
browser-use open https://abc.trycloudflare.com
browser-use state
browser-use screenshot

Form Submission

browser-use open https://example.com/contact
browser-use state
# Shows: [0] input "Name", [1] input "Email", [2] textarea "Message", [3] button "Submit"
browser-use input 0 "John Doe"
browser-use input 1 "john@example.com"
browser-use input 2 "Hello, this is a test message."
browser-use click 3
browser-use state   # Verify success

Screenshot Loop for Visual Verification

browser-use open https://example.com
for i in 1 2 3 4 5; do
  browser-use scroll down
  browser-use screenshot "page_$i.png"
done

Tips

1. Install with --remote-only for sandboxed environments — no --browser flag needed
2. Always run state first to see available elements and their indices
3. Sessions persist across commands — the browser stays open until you close it
4. Tunnels are independent — they don't require or create a browser session, and persist across browser-use close
5. Use --json for programmatic parsing
6. tunnel is idempotent — calling it again for the same port returns the existing URL
7. Close when done — browser-use close closes the browser; browser-use tunnel stop --all stops tunnels

Troubleshooting

"Browser mode 'chromium' not installed"?

• You installed with --remote-only which doesn't include local modes
• This is expected behavior for sandboxed agents
• If you need local browser, reinstall with --full

Cloud browser won't start?

• Verify BROWSER_USE_API_KEY is set
• Check your API key at https://browser-use.com

Tunnel not working?

• Verify cloudflared is installed: which cloudflared
• If missing, install manually (see Setup section) or re-run install.sh --remote-only
• browser-use tunnel list to check active tunnels
• browser-use tunnel stop <port> and retry

Element not found?

• Run browser-use state to see current elements
• browser-use scroll down then browser-use state — element might be below fold
• Page may have changed — re-run state to get fresh indices

Cleanup

Close the browser when done:

browser-use close              # Close browser session
browser-use tunnel stop --all  # Stop all tunnels (if any)

Browser sessions and tunnels are managed separately, so close each as needed.