firefox-browser

name	firefox-browser
description	Control the user's Firefox browser with their logins and cookies intact. Use when you need to browse websites as the user, interact with authenticated pages, fill forms, click buttons, take screenshots, or get page content. (user)
allowed-tools	Bash, Read, Write

Firefox Browser Agent Bridge

Control the user's actual Firefox browser session via WebSocket. This uses their real browser with existing logins and cookies - not a headless browser.

Quick Start

# 0. If Firefox isn't running, start it first
nohup firefox &>/dev/null &

# 1. Check connection
browser ping

# 2. See what tabs are open
browser listTabs '{}'

# 3. Start a new session (recommended)
browser newSession '{"url": "https://example.com"}'

# 4. Read the page with interactable elements marked
browser getContent '{"format": "annotated"}'

Client Usage

browser <action> '<json_params>'

# With timing info
browser --timing <action> '<json_params>'

# With screen recording
browser --record /path/to/dir <action> '<json_params>'

Actions Reference

Session & Tab Management

Action	Description	Key Params
listTabs	List all open tabs across windows	-
newSession	Create new tab to work in	url (optional), sandbox (private window)
setActiveTab	Switch which tab agent works on	tabId, focus
getActiveTab	Get current tab info	-

Navigation & Page Info

Action	Description	Key Params
navigate	Go to URL in current tab	url, wait, newTab
getContent	Get page content	format: annotated, text, html
getInteractables	List clickable elements and inputs	selector (optional scope)
screenshot	Capture visible area as PNG	filename (optional)
reload	Reload current tab	-

Interaction

Action	Description	Key Params
click	Click element	selector, text, or x/y coords
type	Type into focused/selected input	selector, text, submit, clear
fillForm	Fill form fields (inputs, textareas, selects)	fields[] array with selector/value
scroll	Scroll the page or an element	y/x, selector, position
waitFor	Wait for element/text	selector, text, timeout

Rich Text Editors & File Operations

Action	Description	Key Params
uploadFile	Upload file to <input type="file">	selector, path (local file path)
dropFile	Drag-and-drop file onto element	selector, path (local file path)
evaluate	Run JavaScript in page context	script, pageWorld (bool)

uploadFile - Upload Files to Input Elements

browser uploadFile '{"selector": "#fileInput", "path": "/home/user/doc.pdf"}'

The CLI reads the file, base64-encodes it, and sends it via fillForm internally. Works with any <input type="file"> including hidden ones.

dropFile - Drag-and-Drop Files

browser dropFile '{"selector": "#dropZone", "path": "/home/user/image.png"}'

Simulates a native file drop. Creates a DataTransfer with the file and dispatches dragenter → dragover → drop events in the page world. Works with drop zones, contenteditable editors, and any element with drop handlers.

evaluate - Run JavaScript on the Page

# Simple evaluation
browser evaluate '{"script": "return document.title"}'

# In page world (needed for accessing page JS objects)
browser evaluate '{"script": "return window.someAppState", "pageWorld": true}'

Use pageWorld: true when you need to interact with the page's own JavaScript context (React state, app globals, etc). Default runs in content script context.

fillForm - The Right Way to Fill Forms

IMPORTANT: There is no fill command. Use fillForm with a fields array:

# Fill a single field
browser fillForm '{"fields": [{"selector": "#email", "value": "[e-mail]"}]}'

# Fill multiple fields at once (text inputs, textareas, AND select dropdowns)
browser fillForm '{"fields": [
  {"selector": "#name", "value": "John Doe"},
  {"selector": "#email", "value": "[e-mail]"},
  {"selector": "#subject", "value": "support"},
  {"selector": "#message", "value": "Hello world"}
]}'

Works with: <input>, <textarea>, <select>, checkboxes, radio buttons, contenteditable elements, rich text editors (Draft.js, Lexical, TinyMCE, ProseMirror).

For rich text editors, fillForm automatically detects the editor type and uses the appropriate insertion method (execCommand, InputEvent, or direct DOM manipulation in page world).

Control Flow

Action	Description	Key Params
fork	Duplicate tab into multiple paths	paths[] with name + commands
killFork	Close a fork	fork (name)
listForks	List active forks	-
tryUntil	Try alternatives until one succeeds	alternatives[], timeout
parallel	Run commands on multiple URLs	branches[] with url + commands
batch	Run multiple commands in sequence	commands[]

Authentication & Vault

Action	Description	Key Params
autoLogin	Auto-fill credentials from Bitwarden vault and optionally submit	domain, submit (default false)
vaultStatus	Check vault lock state and credential count	-
vaultSync	Re-sync vault from Bitwarden server via API key	-
getAuthContext	Detect login pages, available accounts	-
requestAuth	Request user approval for auth	reason

---

Recommended Workflow

1. Start by Inspecting Available Tabs

browser listTabs '{}'

Returns:

{
  "activeTabId": 123,
  "windows": [
    {
      "windowId": 1,
      "focused": true,
      "tabs": [
        {"tabId": 123, "url": "https://...", "title": "...", "active": true}
      ]
    }
  ],
  "totalTabs": 5
}

2. Start Fresh or Pick Existing Tab

# Start fresh
browser newSession '{"url": "https://amazon.com"}'

# Start in sandbox (private window, no cookies)
browser newSession '{"url": "https://example.com", "sandbox": true}'

# Or switch to existing tab
browser setActiveTab '{"tabId": 456}'

3. Read Page with Annotated Format (Recommended)

browser getContent '{"format": "annotated"}'

Returns content with interactive elements marked inline:

Product Name Here
$4.99
[button: "Add to cart" | selector: #add-btn]
[input:text: "search" | value: "" | selector: #search-box]
[link: "View details" | href: /product/123 | selector: a.details-link]

This shows what's clickable and where it is in context.

4. Handle Login Pages

If you land on a login page or get redirected to one, use autoLogin to fill credentials automatically from the Bitwarden vault:

# Auto-fill credentials and submit the login form
browser autoLogin '{"domain": "github.com", "submit": true}'

This looks up the domain in the vault, fills the username/password, and optionally submits. You don't need to find or interact with form fields manually. After login, use getContent to verify you're logged in.

5. Interact Using Selectors

# Click using selector from annotated output
browser click '{"selector": "#add-btn"}'

# Or by text (prefers visible elements)
browser click '{"text": "Add to cart"}'

# Type into input
browser type '{"selector": "#search-box", "text": "query", "submit": true}'

---

Fork: Speculative Parallel Execution

When you're not sure which path is right, fork the tab and try both:

# Create forks
browser fork '{
  "paths": [
    {
      "name": "google-auth",
      "commands": [{"action": "click", "params": {"text": "Sign in with Google"}}]
    },
    {
      "name": "email-auth",
      "commands": [{"action": "click", "params": {"text": "Sign in with Email"}}]
    }
  ]
}'

Returns:

{
  "forked": true,
  "sourceTabId": 123,
  "forks": [
    {"name": "google-auth", "tabId": 456, "url": "...", "commandResults": [...]},
    {"name": "email-auth", "tabId": 789, "url": "...", "commandResults": [...]}
  ]
}

Work on specific fork:

browser getContent '{"format": "annotated", "fork": "google-auth"}'
browser click '{"text": "Continue", "fork": "google-auth"}'

Kill the wrong path:

browser killFork '{"fork": "email-auth"}'

---

TryUntil: Handle Uncertain UI

When the exact button varies (cookie banners, A/B tests):

browser tryUntil '{
  "alternatives": [
    {"action": "click", "params": {"selector": "#accept-cookies"}},
    {"action": "click", "params": {"text": "Accept All"}},
    {"action": "click", "params": {"selector": ".cookie-dismiss"}}
  ],
  "timeout": 3000
}'

Tries each until one succeeds.

---

Parallel: Multiple URLs at Once

Compare prices across sites:

browser parallel '{
  "branches": [
    {"url": "https://amazon.com/product", "commands": [{"action": "getContent", "params": {"format": "text"}}]},
    {"url": "https://walmart.com/product", "commands": [{"action": "getContent", "params": {"format": "text"}}]}
  ]
}'

---

Authentication (Autonomous Login)

The bridge integrates with a Bitwarden vault (via bronzewarden) for fully autonomous credential fill. No human interaction needed.

Initial Setup

1. Install bronzewarden (local Bitwarden vault client):

cd ~/projects/bronzewarden   # or wherever you cloned it
cargo install --path .

2. Log in to Bitwarden:

# With email + master password
bronzewarden login -e [e-mail]

# Or with API key (if password login is challenged by 2FA/captcha)
# Get your API key from: Bitwarden Web Vault → Settings → Security → Keys → API Key
export BW_CLIENTID="user.xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
export BW_CLIENTSECRET="your_client_secret"
bronzewarden login --apikey

3. Sync the vault:

bronzewarden sync

This downloads and caches your encrypted vault locally at ~/.config/bronzewarden/vault.json.

4. Set up passwordless unlock (so the native host can unlock the vault automatically):

bronzewarden setup-fingerprint

This prompts for your master password one last time, derives the decryption key, and stores it in ~/.config/bronzewarden/protected_key.json. After this, the native host unlocks the vault at startup without any password prompt.

Alternative: password file unlock (if you don't want to use setup-fingerprint):

# Store master password in a file (permissions should be 0600)
echo "your_master_password" > ~/.config/bronzewarden/master_password
chmod 600 ~/.config/bronzewarden/master_password
export BW_PASSWORD_FILE=~/.config/bronzewarden/master_password

Add the BW_PASSWORD_FILE export to native-host-wrapper.sh so it's available when Firefox launches the host.

5. Set up API key for vault sync (needed for browser vaultSync):

# Store your Bitwarden API credentials
echo "user.xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx" > ~/.config/bronzewarden/client_id
echo "your_client_secret" > ~/.config/bronzewarden/client_secret
chmod 600 ~/.config/bronzewarden/client_id ~/.config/bronzewarden/client_secret

Or set environment variables BW_CLIENT_ID and BW_CLIENT_SECRET in native-host-wrapper.sh.

6. Verify it works:

browser ping                    # Check bridge connection
browser vaultStatus '{}'        # Should show: {"locked": false, "loginEntries": N}

Fingerprint Verification (Optional Security)

By default, autoLogin requires fingerprint verification before filling credentials. This is controlled by:

# In native-host-wrapper.sh:
export FAB_AUTOLOGIN_REQUIRE_FINGERPRINT=true   # Require fingerprint touch (default)
export FAB_AUTOLOGIN_REQUIRE_FINGERPRINT=false   # No fingerprint needed (fully autonomous)

When enabled, a desktop notification prompts you to touch the fingerprint sensor before credentials are filled. Requires fprintd to be installed and a fingerprint enrolled.

Auto-Login Flow

# 1. Navigate to the site
browser navigate '{"url": "https://github.com"}'

# 2. Auto-fill credentials (looks up domain in vault, fills form)
browser autoLogin '{"domain": "github.com", "submit": false}'
# Returns: {"filled": true, "maskedUsername": "j***1", "matchedUri": "https://github.com/"}

# 3. Or auto-fill AND submit in one step
browser autoLogin '{"domain": "github.com", "submit": true}'

Vault Management

# Check vault status
browser vaultStatus '{}'
# Returns: {"locked": false, "loginEntries": 322}

# Re-sync vault from server (if credentials were updated)
browser vaultSync '{}'
# Returns: {"synced": true, "loginEntries": 322}

How It Works

• Credentials are stored in Bitwarden and decrypted locally by the native host
• The autoLogin action sends credentials directly to the extension via the native messaging channel (never over WebSocket)
• Vault is auto-unlocked at host startup using a protected key (from setup-fingerprint) or a password file (BW_PASSWORD_FILE)
• Credential lookup priority: BW_PASSWORD env var → BW_PASSWORD_FILE → protected key → gnome-keyring (deprecated)
• API credential lookup priority: BW_CLIENT_ID/BW_CLIENT_SECRET env vars → ~/.config/bronzewarden/client_id/client_secret files

Legacy Auth Detection

# Detect login pages and available accounts
browser getAuthContext '{}'

---

Evaluate: Run JavaScript and Get Results

Execute arbitrary JavaScript in the page context and get the result back:

# Get page title
browser evaluate '{"script": "return document.title"}'
# Returns: {"result": "My Page Title", "type": "string"}

# Count elements
browser evaluate '{"script": "return document.querySelectorAll(\"input\").length"}'
# Returns: {"result": 5, "type": "number"}

# Get form values
browser evaluate '{"script": "return document.querySelector(\"#email\").value"}'
# Returns: {"result": "[e-mail]", "type": "string"}

# Complex queries
browser evaluate '{"script": "return Array.from(document.querySelectorAll(\"input:checked\")).map(el => el.value)"}'
# Returns: {"result": ["option1", "option3"], "type": "object"}

Note: Use return to get a value back. The script runs in page context with full DOM access.

---

Scroll: Navigate Long Pages

Scroll the page by pixels, to elements, or to positions:

# Scroll down 500 pixels
browser scroll '{"y": 500}'

# Scroll up 300 pixels
browser scroll '{"y": -300}'

# Scroll element into view
browser scroll '{"selector": "#section-5"}'

# Scroll to top/bottom
browser scroll '{"position": "top"}'
browser scroll '{"position": "bottom"}'

# Smooth scrolling
browser scroll '{"y": 500, "behavior": "smooth"}'

# Scroll to absolute position
browser scroll '{"scrollTo": {"x": 0, "y": 1000}}'

---

Form State in Annotated Content

The getContent annotated format now shows form element states:

browser getContent '{"format": "annotated"}'

Output includes checked/selected states:

[input:radio: "Option A" | checked: true | selector: #opt-a]
[input:radio: "Option B" | checked: false | selector: #opt-b]
[input:checkbox: "Remember me" | checked: true | selector: #remember]
[select: "Country" | selected: "United States" | selector: #country]
[input:text: "Email" | value: "[e-mail]" | selector: #email]

This is useful for verifying form state without screenshots.

---

Isolated Sessions (for Parallel Execution)

When running multiple tasks in parallel, use tabId to avoid conflicts:

# 1. Create isolated session - get a unique tabId
browser newSession '{"url": "https://example.com"}'
# Returns: {"tabId": 15, "url": "...", "windowId": 1}

# 2. Use that tabId in ALL subsequent commands
browser navigate '{"url": "https://example.com/page", "tabId": 15}'
browser getContent '{"format": "annotated", "tabId": 15}'
browser click '{"selector": "#btn", "tabId": 15}'
browser type '{"selector": "#input", "text": "hello", "tabId": 15}'

This lets multiple agents work in parallel without stepping on each other.

Tips

1. Start with listTabs to see what's open
2. Use newSession for a clean start
3. Use autoLogin when you hit a login page - don't try to fill login forms manually, just browser autoLogin '{"domain": "example.com", "submit": true}'
4. Use tabId for parallel/isolated execution
5. Use annotated format - shows content + clickable elements together
6. Use selectors from annotated output - more reliable than text matching
7. Fork when uncertain - try multiple paths, kill the wrong ones
8. Never use sleep commands - browser commands are synchronous and wait for completion. Use waitFor action if you need to wait for specific elements or text to appear
9. Use uploadFile for file inputs - reads local files and uploads automatically
10. Use dropFile for drop zones - simulates native drag-and-drop
11. Use evaluate for custom JS - with pageWorld: true for page-context access

Troubleshooting

1. Firefox not running? Start it: nohup firefox &>/dev/null &
2. Check connection: browser ping
3. Connection refused? The extension may need to be reloaded in about:debugging
4. Element not found? Use browser getContent '{"format": "annotated"}' to see what's on the page
5. Rich text editor not filling? fillForm handles Draft.js, Lexical, TinyMCE, ProseMirror automatically
6. File drop not working? dropFile runs in page world to bypass Firefox security restrictions