Cxz9909 Agentbrowser

A fast Rust-based headless browser automation CLI with Node.js fallback that enables AI agents to navigate, click, type, and snapshot pages via structured co...

copies: implexa run clawhub/cxz9909-agentbrowser

view source

installs

stars

karma

SkillRank score ↗

7.2/ 10

evaluated by implexa, claude-haiku-4-5 · 2026-05-26

cxz9909-agentbrowser automates web interactions via cli commands for snapshot, click, type, and form submission with ref-based element targeting and session isolation.

structure

7.0

trigger phrases

8.0

procedure

8.0

edge cases

5.0

documentation

8.0

strengths

view original SKILL.md from clawhubclick to expand

---
name: Agent Browser
description: A fast Rust-based headless browser automation CLI with Node.js fallback that enables AI agents to navigate, click, type, and snapshot pages via structured commands.
read_when:
  - Automating web interactions
  - Extracting structured data from pages
  - Filling forms programmatically
  - Testing web UIs
metadata: {"clawdbot":{"emoji":"🌐","requires":{"bins":["node","npm"]}}}
allowed-tools: Bash(agent-browser:*)
---

# Browser Automation with agent-browser

## Installation

### npm recommended

```bash
npm install -g agent-browser
agent-browser install
agent-browser install --with-deps
```

### From Source

```bash
git clone https://github.com/vercel-labs/agent-browser
cd agent-browser
pnpm install
pnpm build
agent-browser install
```

## Quick start

```bash
agent-browser open <url>        # Navigate to page
agent-browser snapshot -i       # Get interactive elements with refs
agent-browser click @e1         # Click element by ref
agent-browser fill @e2 "text"   # Fill input by ref
agent-browser close             # Close browser
```

## Core workflow

1. Navigate: `agent-browser open <url>`
2. Snapshot: `agent-browser snapshot -i` (returns elements with refs like `@e1`, `@e2`)
3. Interact using refs from the snapshot
4. Re-snapshot after navigation or significant DOM changes

## Commands

### Navigation

```bash
agent-browser open <url>      # Navigate to URL
agent-browser back            # Go back
agent-browser forward         # Go forward
agent-browser reload          # Reload page
agent-browser close           # Close browser
```

### Snapshot (page analysis)

```bash
agent-browser snapshot            # Full accessibility tree
agent-browser snapshot -i         # Interactive elements only (recommended)
agent-browser snapshot -c         # Compact output
agent-browser snapshot -d 3       # Limit depth to 3
agent-browser snapshot -s "#main" # Scope to CSS selector
```

### Interactions (use @refs from snapshot)

```bash
agent-browser click @e1           # Click
agent-browser dblclick @e1        # Double-click
agent-browser focus @e1           # Focus element
agent-browser fill @e2 "text"     # Clear and type
agent-browser type @e2 "text"     # Type without clearing
agent-browser press Enter         # Press key
agent-browser press Control+a     # Key combination
agent-browser keydown Shift       # Hold key down
agent-browser keyup Shift         # Release key
agent-browser hover @e1           # Hover
agent-browser check @e1           # Check checkbox
agent-browser uncheck @e1         # Uncheck checkbox
agent-browser select @e1 "value"  # Select dropdown
agent-browser scroll down 500     # Scroll page
agent-browser scrollintoview @e1  # Scroll element into view
agent-browser drag @e1 @e2        # Drag and drop
agent-browser upload @e1 file.pdf # Upload files
```

### Get information

```bash
agent-browser get text @e1        # Get element text
agent-browser get html @e1        # Get innerHTML
agent-browser get value @e1       # Get input value
agent-browser get attr @e1 href   # Get attribute
agent-browser get title           # Get page title
agent-browser get url             # Get current URL
agent-browser get count ".item"   # Count matching elements
agent-browser get box @e1         # Get bounding box
```

### Check state

```bash
agent-browser is visible @e1      # Check if visible
agent-browser is enabled @e1      # Check if enabled
agent-browser is checked @e1      # Check if checked
```

### Screenshots & PDF

```bash
agent-browser screenshot          # Screenshot to stdout
agent-browser screenshot path.png # Save to file
agent-browser screenshot --full   # Full page
agent-browser pdf output.pdf      # Save as PDF
```

### Video recording

```bash
agent-browser record start ./demo.webm    # Start recording (uses current URL + state)
agent-browser click @e1                   # Perform actions
agent-browser record stop                 # Stop and save video
agent-browser record restart ./take2.webm # Stop current + start new recording
```

Recording creates a fresh context but preserves cookies/storage from your session. If no URL is provided, it automatically returns to your current page. For smooth demos, explore first, then start recording.

### Wait

```bash
agent-browser wait @e1                     # Wait for element
agent-browser wait 2000                    # Wait milliseconds
agent-browser wait --text "Success"        # Wait for text
agent-browser wait --url "/dashboard"    # Wait for URL pattern
agent-browser wait --load networkidle      # Wait for network idle
agent-browser wait --fn "window.ready"     # Wait for JS condition
```

### Mouse control

```bash
agent-browser mouse move 100 200      # Move mouse
agent-browser mouse down left         # Press button
agent-browser mouse up left           # Release button
agent-browser mouse wheel 100         # Scroll wheel
```

### Semantic locators (alternative to refs)

```bash
agent-browser find role button click --name "Submit"
agent-browser find text "Sign In" click
agent-browser find label "Email" fill "user@test.com"
agent-browser find first ".item" click
agent-browser find nth 2 "a" text
```

### Browser settings

```bash
agent-browser set viewport 1920 1080      # Set viewport size
agent-browser set device "iPhone 14"      # Emulate device
agent-browser set geo 37.7749 -122.4194   # Set geolocation
agent-browser set offline on              # Toggle offline mode
agent-browser set headers '{"X-Key":"v"}' # Extra HTTP headers
agent-browser set credentials user pass   # HTTP basic auth
agent-browser set media dark              # Emulate color scheme
```

### Cookies & Storage

```bash
agent-browser cookies                     # Get all cookies
agent-browser cookies set name value      # Set cookie
agent-browser cookies clear               # Clear cookies
agent-browser storage local               # Get all localStorage
agent-browser storage local key           # Get specific key
agent-browser storage local set k v       # Set value
agent-browser storage local clear         # Clear all
```

### Network

```bash
agent-browser network route <url>              # Intercept requests
agent-browser network route <url> --abort      # Block requests
agent-browser network route <url> --body '{}'  # Mock response
agent-browser network unroute [url]            # Remove routes
agent-browser network requests                 # View tracked requests
agent-browser network requests --filter api    # Filter requests
```

### Tabs & Windows

```bash
agent-browser tab                 # List tabs
agent-browser tab new [url]       # New tab
agent-browser tab 2               # Switch to tab
agent-browser tab close           # Close tab
agent-browser window new          # New window
```

### Frames

```bash
agent-browser frame "#iframe"     # Switch to iframe
agent-browser frame main          # Back to main frame
```

### Dialogs

```bash
agent-browser dialog accept [text]  # Accept dialog
agent-browser dialog dismiss        # Dismiss dialog
```

### JavaScript

```bash
agent-browser eval "document.title"   # Run JavaScript
```

### State management

```bash
agent-browser state save auth.json    # Save session state
agent-browser state load auth.json    # Load saved state
```

## Example: Form submission

```bash
agent-browser open https://example.com/form
agent-browser snapshot -i
# Output shows: textbox "Email" [ref=e1], textbox "Password" [ref=e2], button "Submit" [ref=e3]

agent-browser fill @e1 "user@example.com"
agent-browser fill @e2 "password123"
agent-browser click @e3
agent-browser wait --load networkidle
agent-browser snapshot -i  # Check result
```

## Example: Authentication with saved state

```bash
# Login once
agent-browser open https://app.example.com/login
agent-browser snapshot -i
agent-browser fill @e1 "username"
agent-browser fill @e2 "password"
agent-browser click @e3
agent-browser wait --url "/dashboard"
agent-browser state save auth.json

# Later sessions: load saved state
agent-browser state load auth.json
agent-browser open https://app.example.com/dashboard
```

## Sessions (parallel browsers)

```bash
agent-browser --session test1 open site-a.com
agent-browser --session test2 open site-b.com
agent-browser session list
```

## JSON output (for parsing)

Add `--json` for machine-readable output:

```bash
agent-browser snapshot -i --json
agent-browser get text @e1 --json
```

## Debugging

```bash
agent-browser open example.com --headed              # Show browser window
agent-browser console                                # View console messages
agent-browser console --clear                        # Clear console
agent-browser errors                                 # View page errors
agent-browser errors --clear                         # Clear errors
agent-browser highlight @e1                          # Highlight element
agent-browser trace start                            # Start recording trace
agent-browser trace stop trace.zip                   # Stop and save trace
agent-browser record start ./debug.webm              # Record from current page
agent-browser record stop                            # Save recording
agent-browser --cdp 9222 snapshot                    # Connect via CDP
```

## Troubleshooting

- If the command is not found on Linux ARM64, use the full path in the bin folder.
- If an element is not found, use snapshot to find the correct ref.
- If the page is not loaded, add a wait command after navigation.
- Use --headed to see the browser window for debugging.

## Options

- --session <name> uses an isolated session.
- --json provides JSON output.
- --full takes a full page screenshot.
- --headed shows the browser window.
- --timeout sets the command timeout in milliseconds.
- --cdp <port> connects via Chrome DevTools Protocol.

## Notes

- Refs are stable per page load but change on navigation.
- Always snapshot after navigation to get new refs.
- Use fill instead of type for input fields to ensure existing text is cleared.

## Reporting Issues

- Skill issues: Open an issue at https://github.com/TheSethRose/Agent-Browser-CLI
- agent-browser CLI issues: Open an issue at https://github.com/vercel-labs/agent-browser

don't have the plugin yet? install it then click "run inline in claude" again.

cxz9909 agentbrowser

Item: Cxz9909 Agentbrowser
Rating: 7.2
Author: Implexa

intent

automate web interactions programmatically via a fast rust-based headless browser cli with node.js fallback. use this skill when you need to navigate pages, extract structured data, fill forms, interact with dynamic content, test uis, or capture page state. the tool exposes an element reference system (@e1, @e2, etc.) that lets agents snapshot a page, locate elements by their accessibility tree, and chain interactions without manual dom inspection.

inputs

required

agent-browser binary: installed globally or available in PATH. install via npm install -g agent-browser or build from source (github.com/vercel-labs/agent-browser).
target URL: full http/https address for navigation commands.

optional context

session name (--session flag): isolate parallel browser contexts. default is unnamed session.
saved state file: json file from prior state save command, contains cookies and local storage.
viewport size: defaults to 1920x1080; customize with set viewport command.
device emulation: preset strings like "iPhone 14" or custom user-agent headers.
http basic auth: username and password for protected resources.
geolocation: latitude, longitude for location-aware testing.
color scheme: "light" or "dark" media preference.
extra http headers: json object of custom request headers.

external connections

chrome devtools protocol (cdp): optional. connect to remote chromium instance via --cdp flag. useful for distributed testing.
network interception: mock, block, or intercept http requests at the CLI level (no external service required, built-in).

edge cases to handle

rate limits: agent-browser itself does not rate-limit, but target websites will. implement delays between requests with wait commands or backoff logic in your agent.
auth expiry: saved state files contain cookies with expiration. re-authenticate if session is stale.
empty snapshots: if snapshot returns no interactive elements, the page may be still loading or use shadow dom. add a wait --load networkidle before retry.
network timeouts: default command timeout is agent-browser's internal value; override with --timeout flag (milliseconds). set realistic values for slow networks.
element refs instability: refs (@e1, @e2) are stable only within the current page load. refs become invalid after navigation or dom reload. always re-snapshot after navigation.
iframe and shadow dom: elements inside iframes require frame switching. shadow dom elements may not appear in accessibility tree; use semantic locators or eval as fallback.
headless detection: some sites block headless browsers. use --headed flag to show browser window or add headers that mask headless state.

procedure

open browser and navigate
- input: target url (string, http/https)
- run: agent-browser open <url>
- output: browser context ready, page loaded or loading
- note: does not wait for network idle by default. use wait --load networkidle after if page is async-heavy.
snapshot page to identify elements
- input: current page state, optional css selector scope (--s flag), optional depth limit (--d flag)
- run: agent-browser snapshot -i (interactive elements only, recommended for agents)
- output: accessibility tree as text, each interactive element labeled with @eN reference (e.g., @e1, @e2, @e3)
- note: use -i flag to filter only clickable, focusable, fillable elements. omit -i for full tree including labels and landmarks.
- json variant: agent-browser snapshot -i --json returns structured json for programmatic parsing.
locate target element by reference
- input: element reference from snapshot output (@e1, @e2, etc.)
- no command needed; reference is valid for steps 4-8 below until next navigation
- output: target element identified in dom
- decision: if element not found in snapshot, element may be hidden, in iframe, or dom may have changed. re-snapshot or use semantic locators (see step 3b).

3b. alternative: semantic locators (if refs unavailable)

input: role, text content, label, or css selector
run: agent-browser find role button click --name "Submit" or agent-browser find text "Sign In" click or agent-browser find label "Email" fill "user@test.com"
output: element located and action applied in one call
note: slower than ref-based workflow but useful for dynamic uis where refs change.

click or focus element
- input: element reference (@eN)
- run: agent-browser click @e1 or agent-browser dblclick @e1 or agent-browser focus @e1
- output: click/focus event fired, page may transition or modal may open
- note: if element is not in viewport, auto-scroll happens. use agent-browser scrollintoview @e1 explicitly if needed.
fill text input
- input: element reference (@eN), text string (quoted if contains spaces)
- run: agent-browser fill @e2 "user@example.com" or agent-browser type @e2 "additional text"
- output: input value set or appended
- decision: use fill to clear existing text first (recommended for forms); use type to append without clearing.
interact with form controls and dropdowns
- input: element reference, optional value string
- run: agent-browser check @e3 or agent-browser uncheck @e3 or agent-browser select @e4 "option-value"
- output: checkbox/radio toggled, dropdown option selected
- note: use agent-browser get value @e4 to retrieve current selected value.
press keys and keyboard shortcuts
- input: key name or combination (e.g., "Enter", "Control+a", "Shift+Tab")
- run: agent-browser press Enter or agent-browser press Control+a
- output: keyboard event fired
- variants: keydown <key> holds key, keyup <key> releases, useful for modifier sequences.
wait for page changes and network events
- input: wait target (element ref, css selector, text, url pattern, network condition, js condition, or milliseconds)
- run: agent-browser wait @e1 or agent-browser wait 2000 or agent-browser wait --text "Success" or agent-browser wait --load networkidle or agent-browser wait --fn "window.ready"
- output: wait completes when condition met or timeout exceeded
- note: use --load networkidle after form submission or navigation to ensure async updates finish.
extract page data
- input: element reference or page level target
- run: agent-browser get text @e1 or agent-browser get html @e1 or agent-browser get value @e1 or agent-browser get attr @e1 href or agent-browser get title or agent-browser get url
- output: string (text, html, attribute value) or page-level metadata
- json variant: agent-browser get text @e1 --json for machine parsing.
- note: returns null or empty string if attribute/text not present.
check element state
- input: element reference, state type (visible, enabled, checked)
- run: agent-browser is visible @e1 or agent-browser is enabled @e1 or agent-browser is checked @e1
- output: boolean or state string
- use case: conditional logic in agent workflows (if visible, click; else retry).
scroll and mouse control
- input: direction/distance or coordinates
- run: agent-browser scroll down 500 or agent-browser scrollintoview @e1 or agent-browser mouse move 100 200 or agent-browser mouse down left or agent-browser mouse wheel 100
- output: page scrolled or mouse pointer moved
- note: drag-and-drop via agent-browser drag @e1 @e2.
upload files
- input: element reference (file input), local file path
- run: agent-browser upload @e1 /path/to/file.pdf
- output: file uploaded to input element
- note: file path must be local to the machine running agent-browser. supports multiple files if input allows.
capture page state
- input: optional output file path, optional --full flag for full-page
- run: agent-browser screenshot (stdout) or agent-browser screenshot path.png or agent-browser screenshot --full
- output: png image bytes or file written
- pdf variant: agent-browser pdf output.pdf generates pdf.
record video
- input: output file path (.webm), current page url (preserved in recording context)
- run: agent-browser record start ./demo.webm (start recording), perform actions, agent-browser record stop (save)
- output: .webm video file containing browser interactions
- note: recording creates isolated context but preserves cookies/storage. if no url given at start, returns to current page.
manage cookies and storage
- input: action type (get/set/clear), optional key/value
- run: agent-browser cookies (list all) or agent-browser cookies set name value or agent-browser cookies clear
- output: cookie list (json or text) or operation confirmation
- variants: agent-browser storage local (list), agent-browser storage local set k v (set), agent-browser storage local clear (clear).
intercept and mock network requests
- input: url pattern, optional --abort flag (block), optional --body flag (mock response json)
- run: agent-browser network route "https://api.example.com/*" or `agent-

Cxz9909 Agentbrowser

related skills

cxz9909 agentbrowser

intent

inputs

required

optional context

external connections

edge cases to handle

procedure