MCP Tools Reference¶

Complete reference for all 169 MCP tools across 24 namespaces.

Naming Convention¶

All tool names follow the pattern: {namespace}_{verb}_{target}

Naming Principles:

Explicit verbs: element_get_text not element_text
No abbreviations: wait_for_function not wait_fn
Full words: human_pause not hitl_pause, accessibility_snapshot not a11y_snapshot
Consistent verb patterns: get/set, start/stop, is/has

Namespace	Purpose	Count
`accessibility_`	Accessibility tree	1
`batch_`	Multi-step batch execution	1
`browser_`	Browser lifecycle	2
`cdp_`	Chrome DevTools Protocol	20
`config_`	Configuration	1
`console_`	Console messages	2
`dialog_`	Dialog handling	2
`element_`	Element interactions and state	33
`frame_`	Frame selection	2
`http_`	HTTP requests in browser context	1
`human_`	Human-in-the-loop	1
`input_`	Low-level keyboard/mouse/touch	12
`js_`	JavaScript execution	4
`network_`	Network requests and mocking	6
`page_`	Page navigation, state, screenshots, emulation	20
`record_`	Script recording	5
`state_`	Named state snapshots	4
`storage_`	Cookies, localStorage, sessionStorage	17
`tab_`	Tab management	3
`test_`	Assertions, verification, reporting	16
`trace_`	Tracing	6
`video_`	Video recording	2
`wait_`	Waiting operations	6
`workflow_`	High-level automation workflows	2

Machine-Readable Format¶

For programmatic access, use:

# Export as JSON
w3pilot mcp --list-tools > mcp-tools.json

# Or via the standalone binary
w3pilot-mcp --list-tools

The JSON format follows the MCP protocol structure:

{
  "tools": [
    {
      "name": "tool_name",
      "description": "Tool description",
      "category": "Category Name"
    }
  ],
  "categories": {
    "Category Name": 5
  },
  "total": 169
}

A pre-generated version is available at mcp-tools.json.

Browser Management¶

browser_launch¶

Launch a browser instance.

Input:

Field	Type	Description
`headless`	boolean	Run without GUI

Output:

Field	Type	Description
`message`	string	Status message

browser_quit¶

Close the browser.

Output:

Field	Type	Description
`message`	string	Status message

page_navigate¶

Navigate to a URL.

Input:

Field	Type	Required	Description
`url`	string	✅	Target URL

Output:

Field	Type	Description
`url`	string	Current URL
`title`	string	Page title

page_go_back¶

Navigate back in history.

page_go_forward¶

Navigate forward in history.

page_reload¶

Reload the current page.

page_scroll¶

Scroll the page or a specific element.

Input:

Field	Type	Required	Description
`direction`	string	✅	Scroll direction: up, down, left, right
`amount`	integer		Pixels to scroll (0 for full page)
`selector`	string		CSS selector to scroll within

Output:

Field	Type	Description
`message`	string	Status message

Element Interactions¶

element_click¶

Click an element.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`timeout_ms`	integer		Timeout (default: 5000)

element_double_click¶

Double-click an element.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`timeout_ms`	integer		Timeout

element_type¶

Type text into an element (appends).

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`text`	string	✅	Text to type
`timeout_ms`	integer		Timeout

element_fill¶

Fill an input (replaces content).

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`value`	string	✅	Value to fill
`timeout_ms`	integer		Timeout

element_fill_form¶

Fill multiple form fields at once.

Input:

Field	Type	Required	Description
`fields`	array	✅	Array of {selector, value} objects
`timeout_ms`	integer		Timeout per field

Output:

Field	Type	Description
`message`	string	Status message
`filled`	integer	Number of fields filled
`errors`	array	Any errors encountered

element_clear¶

Clear an input element.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector

element_press¶

Press a key on an element.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`key`	string	✅	Key (e.g., "Enter")

Form Controls¶

element_check¶

Check a checkbox.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector

element_uncheck¶

Uncheck a checkbox.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector

element_select¶

Select dropdown option(s).

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`values`	array		Option values
`labels`	array		Option labels
`indexes`	array		Option indexes

element_set_files¶

Set files on a file input.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`files`	array	✅	File paths

Element State¶

element_get_text¶

Get element text content.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector

Output:

Field	Type	Description
`text`	string	Text content

element_get_value¶

Get input element value.

element_get_inner_html¶

Get element innerHTML.

element_get_outer_html¶

Get element outerHTML (including the element itself).

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector

Output:

Field	Type	Description
`html`	string	Element's outer HTML

element_get_inner_text¶

Get element innerText.

element_get_attribute¶

Get element attribute.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`name`	string	✅	Attribute name

element_get_bounding_box¶

Get element bounding box.

Output:

Field	Type	Description
`x`	number	X position
`y`	number	Y position
`width`	number	Width
`height`	number	Height

element_is_visible¶

Check if element is visible.

Output:

Field	Type	Description
`visible`	boolean	Visibility state

element_is_hidden¶

Check if element is hidden.

element_is_enabled¶

Check if element is enabled.

element_is_checked¶

Check if checkbox/radio is checked.

element_is_editable¶

Check if element is editable.

element_get_role¶

Get ARIA role.

element_get_label¶

Get accessible label.

Page State¶

page_get_title¶

Get page title.

Output:

Field	Type	Description
`title`	string	Page title

page_get_url¶

Get current URL.

Output:

Field	Type	Description
`url`	string	Current URL

page_get_content¶

Get page HTML content.

page_set_content¶

Set page HTML content.

page_get_viewport¶

Get viewport dimensions.

page_set_viewport¶

Set viewport dimensions.

Input:

Field	Type	Required	Description
`width`	integer	✅	Width
`height`	integer	✅	Height

Screenshots & PDF¶

page_screenshot¶

Capture page screenshot.

Input:

Field	Type	Description
`format`	string	"base64" or "file"

Output:

Field	Type	Description
`data`	string	Base64 image data

element_screenshot¶

Capture element screenshot.

page_pdf¶

Generate PDF.

JavaScript¶

js_evaluate¶

Execute JavaScript.

Input:

Field	Type	Required	Description
`script`	string	✅	JavaScript code
`max_result_size`	integer		Truncate result to this many characters (0 = no truncation)

Output:

Field	Type	Description
`result`	any	Evaluation result
`truncated`	boolean	Whether result was truncated

element_evaluate¶

Evaluate JavaScript with element.

js_add_script¶

Inject JavaScript.

js_add_style¶

Inject CSS.

HTTP Requests¶

http_request¶

Make HTTP requests in the browser context with session cookies.

Input:

Field	Type	Required	Description
`url`	string	✅	Target URL
`method`	string		HTTP method (default: GET)
`headers`	object		Request headers
`body`	string		Request body for POST/PUT/PATCH
`content_type`	string		Content-Type header
`max_body_length`	integer		Truncate response body (default: 8192)

Output:

Field	Type	Description
`status`	integer	HTTP status code
`statusText`	string	HTTP status text
`headers`	object	Response headers
`body`	string	Response body
`truncated`	boolean	Whether body was truncated
`url`	string	Final URL (after redirects)

Batch Execution¶

batch_execute¶

Execute multiple operations in a single MCP call to reduce round-trip latency.

Input:

Field	Type	Required	Description
`steps`	array	✅	Array of tool steps to execute
`stop_on_error`	boolean		Stop on first error (default: true)
`continue_on_error`	boolean		Continue even if steps fail

Each step object:

Field	Type	Required	Description
`tool`	string	✅	Tool name to execute
`args`	object		Arguments for the tool

Output:

Field	Type	Description
`results`	array	Results for each step
`total_steps`	integer	Total steps in batch
`success_count`	integer	Successful steps
`failure_count`	integer	Failed steps
`stopped_early`	boolean	Whether stopped before completion
`total_duration_ms`	integer	Total execution time

Supported Tools:

Category	Tools
Navigation	`page_navigate`, `page_go_back`, `page_go_forward`, `page_reload`
Page Info	`page_get_title`, `page_get_url`, `page_screenshot`
Elements	`element_click`, `element_fill`, `element_type`, `element_get_text`
JavaScript	`js_evaluate`
Waiting	`wait_for_selector`, `wait_for_load`
HTTP	`http_request`

Example:

{
  "tool": "batch_execute",
  "arguments": {
    "steps": [
      {"tool": "page_navigate", "args": {"url": "https://example.com/login"}},
      {"tool": "element_fill", "args": {"selector": "#username", "value": "user"}},
      {"tool": "element_fill", "args": {"selector": "#password", "value": "pass"}},
      {"tool": "element_click", "args": {"selector": "#login"}},
      {"tool": "wait_for_selector", "args": {"selector": "#dashboard"}}
    ]
  }
}

Waiting¶

wait_for_state¶

Wait for element state.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector
`state`	string	✅	visible/hidden/attached/detached

wait_for_url¶

Wait for URL pattern.

Input:

Field	Type	Required	Description
`pattern`	string	✅	URL pattern

wait_for_load¶

Wait for load state.

Input:

Field	Type	Required	Description
`state`	string	✅	load/domcontentloaded/networkidle

wait_for_function¶

Wait for JavaScript function.

Input Controllers¶

input_keyboard_press¶

Press a key.

input_keyboard_down¶

Hold a key.

input_keyboard_up¶

Release a key.

input_keyboard_type¶

Type text via keyboard.

input_mouse_click¶

Click at coordinates.

input_mouse_move¶

Move mouse.

input_mouse_down¶

Press mouse button.

input_mouse_up¶

Release mouse button.

input_mouse_wheel¶

Scroll mouse wheel.

input_touch_tap¶

Tap at coordinates.

input_touch_swipe¶

Swipe gesture.

input_mouse_drag¶

Drag from one point to another using the mouse.

Input:

Field	Type	Required	Description
`start_x`	number	✅	Starting X coordinate
`start_y`	number	✅	Starting Y coordinate
`end_x`	number	✅	Ending X coordinate
`end_y`	number	✅	Ending Y coordinate
`steps`	integer		Number of intermediate steps (default: 10)

Page Management¶

page_new¶

Create new page/tab.

page_get_count¶

Get page count.

page_close¶

Close current page.

page_bring_to_front¶

Activate page.

Emulation¶

page_emulate_media¶

Emulate CSS media features for accessibility testing.

Input:

Field	Type	Description
`media`	string	Media type: "screen" or "print"
`color_scheme`	string	Color scheme: "light", "dark", "no-preference"
`reduced_motion`	string	Reduced motion: "reduce", "no-preference"
`forced_colors`	string	Forced colors: "active", "none"
`contrast`	string	Contrast: "more", "less", "no-preference"

Output:

Field	Type	Description
`message`	string	Status message
`settings`	array	Applied settings

Accessibility Testing Use Cases:

color_scheme: Test dark/light mode support for users with light sensitivity
reduced_motion: Test that animations are disabled for users with vestibular disorders
forced_colors: Test Windows High Contrast Mode compatibility
contrast: Test increased/decreased contrast for users with low vision

page_set_geolocation¶

Set the browser's geolocation.

Input:

Field	Type	Required	Description
`latitude`	number	Yes	Latitude coordinate
`longitude`	number	Yes	Longitude coordinate
`accuracy`	number		Accuracy in meters

Tab Management¶

tab_list¶

List all open browser tabs.

Output:

Field	Type	Description
`tabs`	array	Tab information (index, id, url, title)
`count`	integer	Number of open tabs
`current_tab`	integer	Index of the current tab

tab_select¶

Switch to a specific tab.

Input:

Field	Type	Required	Description
`index`	integer		Tab index (0-based)
`id`	string		Tab ID from tab_list

tab_close¶

Close a specific tab.

Input:

Field	Type	Required	Description
`index`	integer		Tab index to close
`id`	string		Tab ID to close

Dialog Handling¶

dialog_handle¶

Handle a browser dialog (alert, confirm, prompt, beforeunload).

Input:

Field	Type	Required	Description
`action`	string	Yes	Action: "accept" or "dismiss"
`prompt_text`	string		Text for prompt dialogs

dialog_get¶

Get information about the current dialog.

Output:

Field	Type	Description
`has_dialog`	boolean	Whether a dialog is open
`dialog_type`	string	alert, confirm, prompt, beforeunload
`message`	string	Dialog message
`default_value`	string	Default value for prompt dialogs

Console Messages¶

console_get_messages¶

Get console messages from the page.

Input:

Field	Type	Required	Description
`level`	string		Filter by level: log, info, warn, error, debug
`clear`	boolean		Clear messages after retrieving

Output:

Field	Type	Description
`messages`	array	Console message objects
`count`	integer	Number of messages

Each message contains:

Field	Type	Description
`type`	string	Message type (log, info, warn, error, debug)
`text`	string	Message text
`args`	array	Additional arguments
`url`	string	Source URL
`line`	integer	Source line number

console_clear¶

Clear all buffered console messages.

Network Requests¶

network_get_requests¶

Get captured network requests.

Input:

Field	Type	Description
`url_pattern`	string	Filter by URL pattern (glob or regex)
`method`	string	Filter by HTTP method
`resource_type`	string	Filter by resource type
`clear`	boolean	Clear requests after retrieving

Output:

Field	Type	Description
`requests`	array	Network request objects
`count`	integer	Number of requests

Each request contains:

Field	Type	Description
`url`	string	Request URL
`method`	string	HTTP method
`resource_type`	string	Resource type
`status`	integer	Response status code
`status_text`	string	Response status text
`response_size`	integer	Response size in bytes

network_clear¶

Clear all buffered network requests.

Network Mocking¶

network_route¶

Register a mock response for requests matching a URL pattern.

Input:

Field	Type	Required	Description
`pattern`	string	✅	URL pattern (glob or regex, e.g., `*/api/`)
`status`	integer		HTTP status code (default: 200)
`body`	string		Response body content
`content_type`	string		Content-Type header (default: application/json)
`headers`	object		Additional response headers

Output:

Field	Type	Description
`message`	string	Status message
`pattern`	string	Registered pattern

network_list_routes¶

List all active route handlers.

Output:

Field	Type	Description
`routes`	array	Route info objects
`count`	integer	Number of active routes

Each route contains:

Field	Type	Description
`pattern`	string	URL pattern
`status`	integer	Response status code
`content_type`	string	Content-Type header

network_unroute¶

Remove a previously registered route handler.

Input:

Field	Type	Required	Description
`pattern`	string	✅	URL pattern to unregister

Output:

Field	Type	Description
`message`	string	Status message

network_set_offline¶

Set the browser's network state for offline mode testing.

Input:

Field	Type	Required	Description
`offline`	boolean	✅	Set to true for offline mode

Output:

Field	Type	Description
`message`	string	Status message
`offline`	boolean	Current offline state

Cookies & Storage¶

storage_get_cookies¶

Get browser cookies.

storage_set_cookies¶

Set browser cookies.

storage_clear_cookies¶

Clear all cookies.

storage_get_state¶

Get complete browser storage state including cookies, localStorage, and sessionStorage as JSON. This can be saved to a file and later restored using storage_set_state to resume a session.

Output:

Returns JSON containing:

cookies: Array of all cookies
origins: Array of origin storage, each with:
origin: The origin URL (e.g., "https://example.com")
localStorage: Key-value map of localStorage items
sessionStorage: Key-value map of sessionStorage items (for current page's origin)

storage_set_state¶

Restore browser storage from JSON (output of storage_get_state). Restores cookies, localStorage, and sessionStorage.

Input:

Field	Type	Required	Description
`state`	string	Yes	JSON from storage_get_state

storage_clear_all¶

Clear all browser storage including cookies, localStorage, and sessionStorage.

LocalStorage¶

storage_local_get¶

Get a value from localStorage by key.

Input:

Field	Type	Required	Description
`key`	string	Yes	Key to retrieve

Output:

Field	Type	Description
`key`	string	The key
`value`	string	The value (null if not found)

storage_local_set¶

Set a value in localStorage.

Input:

Field	Type	Required	Description
`key`	string	Yes	Key to set
`value`	string	Yes	Value to store

storage_local_delete¶

Delete a key from localStorage.

Input:

Field	Type	Required	Description
`key`	string	Yes	Key to delete

storage_local_clear¶

Clear all localStorage data for the current origin.

storage_local_list¶

List all keys and values in localStorage.

Output:

Field	Type	Description
`items`	object	Key-value pairs
`count`	integer	Number of items

SessionStorage¶

storage_session_get¶

Get a value from sessionStorage by key.

Input:

Field	Type	Required	Description
`key`	string	Yes	Key to retrieve

Output:

Field	Type	Description
`key`	string	The key
`value`	string	The value (null if not found)

storage_session_set¶

Set a value in sessionStorage.

Input:

Field	Type	Required	Description
`key`	string	Yes	Key to set
`value`	string	Yes	Value to store

storage_session_delete¶

Delete a key from sessionStorage.

Input:

Field	Type	Required	Description
`key`	string	Yes	Key to delete

storage_session_clear¶

Clear all sessionStorage data for the current origin.

storage_session_list¶

List all keys and values in sessionStorage.

Output:

Field	Type	Description
`items`	object	Key-value pairs
`count`	integer	Number of items

Script Recording¶

record_start¶

Begin recording actions.

Input:

Field	Type	Description
`name`	string	Script name
`description`	string	Description
`baseUrl`	string	Base URL

record_stop¶

Stop recording.

Output:

Field	Type	Description
`stepCount`	integer	Steps recorded

record_export¶

Export recorded script.

Output:

Field	Type	Description
`script`	string	JSON script
`stepCount`	integer	Steps
`format`	string	Output format

record_get_status¶

Check recording state.

record_clear¶

Clear recorded steps.

Tracing¶

Tools for recording detailed execution traces for debugging and analysis. Traces include screenshots, DOM snapshots, and network activity. View traces with npx playwright show-trace <trace.zip>.

trace_start¶

Start trace recording with screenshots and DOM snapshots.

Input:

Field	Type	Description
`name`	string	Trace name (used for file naming)
`title`	string	Title shown in trace viewer
`screenshots`	boolean	Include screenshots (default: true)
`snapshots`	boolean	Include DOM snapshots (default: true)
`sources`	boolean	Include source files

Output:

Field	Type	Description
`message`	string	Status message
`name`	string	Trace name

trace_stop¶

Stop trace recording and save or return the trace data.

Input:

Field	Type	Required	Description
`path`	string		File path to save trace ZIP (returns base64 if omitted)

Output:

Field	Type	Description
`message`	string	Status message
`path`	string	File path if saved
`data`	string	Base64-encoded ZIP if no path specified
`size_kb`	integer	Trace size in KB
`view_hint`	string	Command to view trace

trace_chunk_start¶

Start a new trace chunk within an active trace for segmenting recordings.

Input:

Field	Type	Required	Description
`name`	string		Chunk name
`title`	string		Chunk title shown in trace viewer

trace_chunk_stop¶

Stop the current trace chunk.

Input:

Field	Type	Required	Description
`path`	string		File path to save chunk ZIP

Output:

Field	Type	Description
`message`	string	Status message
`path`	string	File path if saved
`data`	string	Base64-encoded ZIP if no path
`size_kb`	integer	Chunk size in KB

trace_group_start¶

Start a trace group for logical grouping of actions in the trace viewer.

Input:

Field	Type	Required	Description
`name`	string	✅	Group name
`location`	string		Source location to associate

trace_group_stop¶

Stop the current trace group.

Init Scripts¶

js_init_script¶

Add JavaScript that runs before page scripts on every navigation. Useful for mocking APIs, injecting test helpers, or setting up authentication.

Input:

Field	Type	Required	Description
`script`	string	✅	JavaScript code to inject before page scripts

Output:

Field	Type	Description
`message`	string	Status message

Example Use Cases:

// Mock fetch API
window.fetch = async (url) => {
  if (url.includes('/api/user')) {
    return { json: () => ({ id: 1, name: 'Test User' }) };
  }
  return originalFetch(url);
};

// Disable analytics
window.gtag = () => {};
window.analytics = { track: () => {}, identify: () => {} };

// Inject test helpers
window.testHelpers = {
  fillForm: (data) => { /* ... */ },
  waitForElement: (sel) => { /* ... */ }
};

Assertions¶

test_assert_text¶

Assert text exists.

Input:

Field	Type	Required	Description
`text`	string	✅	Expected text
`selector`	string		Limit to element

test_assert_element¶

Assert element exists.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector

Testing Tools¶

test_verify_value¶

Verify that an input element has the expected value.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector for the input element
`expected`	string	✅	Expected value to verify
`timeout_ms`	integer		Timeout in milliseconds (default: 5000)

Output:

Field	Type	Description
`passed`	boolean	Whether the verification passed
`actual`	string	The actual value found
`message`	string	Status message

test_verify_list¶

Verify that a list of text items are all visible on the page.

Input:

Field	Type	Required	Description
`items`	array	✅	List of text items that should be visible
`selector`	string		Optional CSS selector to scope the search
`timeout_ms`	integer		Timeout in milliseconds (default: 5000)

Output:

Field	Type	Description
`passed`	boolean	Whether all items were found
`found`	array	Items that were found
`missing`	array	Items that were not found
`message`	string	Status message

test_generate_locator¶

Generate a locator string for a given element using a specific strategy.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector for the element
`strategy`	string		Locator strategy: css, xpath, testid, role, text (default: css)
`timeout_ms`	integer		Timeout in milliseconds (default: 5000)

Output:

Field	Type	Description
`locator`	string	Generated locator string
`strategy`	string	Strategy used
`metadata`	object	Additional metadata (role, label, text, etc.)

Test Reporting¶

test_get_report¶

Get test execution report.

test_reset¶

Clear test results.

test_set_target¶

Set test target description.

State Management¶

Named state snapshots for saving and restoring browser state across sessions.

state_save¶

Save current browser state (cookies, localStorage, sessionStorage) to a named snapshot.

Input:

Field	Type	Required	Description
`name`	string	✅	Name for the state snapshot

Output:

Field	Type	Description
`message`	string	Status message
`path`	string	Path to saved state file

state_load¶

Load a previously saved state snapshot.

Input:

Field	Type	Required	Description
`name`	string	✅	Name of the state snapshot to load

Output:

Field	Type	Description
`message`	string	Status message

state_list¶

List all saved state snapshots.

Output:

Field	Type	Description
`states`	array	List of state names
`count`	integer	Number of saved states

state_delete¶

Delete a saved state snapshot.

Input:

Field	Type	Required	Description
`name`	string	✅	Name of the state snapshot to delete

Output:

Field	Type	Description
`message`	string	Status message

Workflow Automation¶

High-level workflow tools for common automation patterns.

workflow_login¶

Automated login workflow with credential input and success verification.

Input:

Field	Type	Required	Description
`url`	string	✅	Login page URL
`username_selector`	string	✅	CSS selector for username field
`password_selector`	string	✅	CSS selector for password field
`submit_selector`	string	✅	CSS selector for submit button
`username`	string	✅	Username to enter
`password`	string	✅	Password to enter
`success_url`	string		Expected URL after successful login
`success_selector`	string		Element that indicates successful login

Output:

Field	Type	Description
`success`	boolean	Whether login succeeded
`url`	string	Final URL after login
`message`	string	Status message

workflow_extract_table¶

Extract data from an HTML table to structured JSON.

Input:

Field	Type	Required	Description
`selector`	string	✅	CSS selector for the table
`headers`	boolean		Whether first row contains headers (default: true)

Output:

Field	Type	Description
`rows`	array	Array of row objects
`columns`	array	Column names
`count`	integer	Number of rows

Configuration¶

config_get¶

Get the resolved MCP server configuration.

Output:

Field	Type	Description
`headless`	boolean	Whether browser runs headless
`project`	string	Project name for reports
`default_timeout_ms`	integer	Default timeout in milliseconds
`browser_launched`	boolean	Whether browser has been launched

CDP Tools (Chrome DevTools Protocol)¶

These tools use direct CDP connection for advanced browser profiling and debugging that isn't available through WebDriver BiDi.

Performance & Profiling¶

cdp_get_performance_metrics¶

Get Core Web Vitals (LCP, CLS, INP) and navigation timing metrics.

Output:

Field	Type	Description
`lcp`	object	Largest Contentful Paint data
`cls`	number	Cumulative Layout Shift score
`inp`	object	Interaction to Next Paint data
`navigation`	object	Navigation timing metrics

cdp_get_memory_stats¶

Get JavaScript heap memory statistics.

Output:

Field	Type	Description
`usedJSHeapSize`	integer	Used heap size in bytes
`totalJSHeapSize`	integer	Total heap size in bytes
`jsHeapSizeLimit`	integer	Maximum heap size in bytes

cdp_take_heap_snapshot¶

Capture a V8 heap snapshot for memory profiling.

Input:

Field	Type	Required	Description
`path`	string		File path to save snapshot (default: auto-generated)

Output:

Field	Type	Description
`path`	string	Path to saved .heapsnapshot file
`size`	integer	Snapshot size in bytes

Network Emulation (CDP)¶

cdp_emulate_network¶

Emulate network conditions.

Input:

Field	Type	Description
`preset`	string	Preset: "slow3g", "fast3g", "4g", "wifi", "offline"
`latency`	number	Custom latency in ms
`download`	number	Custom download throughput in bytes/sec
`upload`	number	Custom upload throughput in bytes/sec

cdp_emulate_cpu¶

Emulate CPU throttling.

Input:

Field	Type	Required	Description
`rate`	number	✅	Throttling rate (2, 4, or 6 for 2x, 4x, 6x slowdown)

cdp_clear_network_emulation¶

Clear network emulation and restore normal conditions.

cdp_clear_cpu_emulation¶

Clear CPU emulation and restore normal speed.

Quality Auditing¶

cdp_run_lighthouse¶

Run Lighthouse audit for performance, accessibility, SEO, and best practices.

Input:

Field	Type	Description
`categories`	array	Categories: "performance", "accessibility", "seo", "best-practices"
`device`	string	Device: "desktop" or "mobile" (default: desktop)
`output_dir`	string	Directory for HTML/JSON reports

Output:

Field	Type	Description
`url`	string	Audited URL
`scores`	object	Scores by category (0-100)
`passed_audits`	integer	Number of passing audits
`failed_audits`	integer	Number of failing audits
`report_paths`	object	Paths to generated reports

Code Coverage (CDP)¶

cdp_start_coverage¶

Start collecting JavaScript and CSS code coverage.

Input:

Field	Type	Description
`js`	boolean	Include JavaScript coverage (default: true)
`css`	boolean	Include CSS coverage (default: true)
`call_count`	boolean	Include call counts for JS functions
`detailed`	boolean	Include detailed block coverage

cdp_stop_coverage¶

Stop coverage collection and return the coverage report.

Output:

Field	Type	Description
`js`	object	JavaScript coverage summary
`css`	object	CSS coverage summary
`scripts`	array	Detailed script coverage

Console Debugging (CDP)¶

cdp_enable_console_debugger¶

Enable enhanced console debugging with stack traces.

cdp_get_console_entries¶

Get console messages with full stack traces.

Output:

Field	Type	Description
`entries`	array	Console entries with type, args, stack trace
`count`	integer	Number of entries

cdp_get_browser_logs¶

Get browser logs including deprecations, interventions, and violations.

Output:

Field	Type	Description
`logs`	array	Log entries with source, level, text
`count`	integer	Number of logs

cdp_disable_console_debugger¶

Disable enhanced console debugging.

Screencast (CDP)¶

cdp_start_screencast¶

Start streaming screen frames.

Input:

Field	Type	Description
`format`	string	Image format: "jpeg" or "png" (default: jpeg)
`quality`	integer	Image quality 0-100 (default: 80)
`max_width`	integer	Maximum width in pixels
`max_height`	integer	Maximum height in pixels

cdp_stop_screencast¶

Stop screen streaming.

Extensions Management (CDP)¶

cdp_install_extension¶

Install a Chrome extension from a local path.

Input:

Field	Type	Required	Description
`path`	string	✅	Path to unpacked extension directory

Output:

Field	Type	Description
`id`	string	Installed extension ID

cdp_uninstall_extension¶

Uninstall a Chrome extension by ID.

Input:

Field	Type	Required	Description
`id`	string	✅	Extension ID to uninstall

cdp_list_extensions¶

List all installed Chrome extensions.

Output:

Field	Type	Description
`extensions`	array	Extension info (id, name, enabled)
`count`	integer	Number of extensions

Network Request Bodies (CDP)¶

cdp_get_response_body¶

Get the response body for a specific network request by ID.

Input:

Field	Type	Required	Description
`request_id`	string	✅	Network request ID from network_get_requests
`save_to_file`	string		Optional file path to save binary content

Output:

Field	Type	Description
`body`	string	Response body content
`base64_encoded`	boolean	Whether body is base64 encoded
`path`	string	File path if saved to file

MCP Tools Reference¶

Naming Convention¶

Machine-Readable Format¶

Browser Management¶

browser_launch¶

browser_quit¶

Navigation¶

page_navigate¶

page_go_back¶

page_go_forward¶

page_reload¶

page_scroll¶

Element Interactions¶

element_click¶

element_double_click¶

element_type¶

element_fill¶

element_fill_form¶

element_clear¶

element_press¶

Form Controls¶

element_check¶

element_uncheck¶

element_select¶

element_set_files¶

Element State¶

element_get_text¶

element_get_value¶

element_get_inner_html¶

element_get_outer_html¶

element_get_inner_text¶

element_get_attribute¶

element_get_bounding_box¶

element_is_visible¶

element_is_hidden¶

element_is_enabled¶

element_is_checked¶

element_is_editable¶

element_get_role¶

element_get_label¶

Page State¶

page_get_title¶

page_get_url¶

page_get_content¶

page_set_content¶

page_get_viewport¶

page_set_viewport¶

Screenshots & PDF¶

page_screenshot¶

element_screenshot¶

page_pdf¶

JavaScript¶

js_evaluate¶

element_evaluate¶

js_add_script¶

js_add_style¶

HTTP Requests¶

http_request¶

Batch Execution¶

batch_execute¶

Waiting¶

wait_for_state¶

wait_for_url¶

wait_for_load¶

wait_for_function¶

Input Controllers¶

input_keyboard_press¶

input_keyboard_down¶

input_keyboard_up¶

input_keyboard_type¶

input_mouse_click¶

input_mouse_move¶

input_mouse_down¶

input_mouse_up¶

input_mouse_wheel¶

input_touch_tap¶

input_touch_swipe¶

input_mouse_drag¶

Page Management¶

page_new¶