Get context information about the current MCP tab group. Returns all tab IDs inside the group if it exists. CRITICAL: You must get the context at least once before using other browser automation tools so you know what tabs exist. Each new conversation should create its own new tab (using tabs_create) rather than reusing existing tabs, unless the user explicitly asks to use an existing tab.

Creates a new empty tab in the MCP tab group.

Navigate to a URL, or go forward/back in browser history. If you don't have a valid tab ID, use tabs_context first to get available tabs.

Use a mouse and keyboard to interact with a web browser, and take screenshots. If you don't have a valid tab ID, use tabs_context first to get available tabs.

• Whenever you intend to click on an element like an icon, you should consult a screenshot to determine the coordinates of the element before moving the cursor.

• If you tried clicking on a program or link but it failed to load, even after waiting, try adjusting your click location so that the tip of the cursor visually falls on the element that you want to click.

• Make sure to click any buttons, links, icons, etc with the cursor tip in the center of the element. Don't click boxes on their edges unless asked.

Find elements on the page using natural language. Can search for elements by their purpose (e.g., "search bar", "login button") or by text content (e.g., "organic mango product"). Returns up to 20 matching elements with references that can be used with other tools. If more than 20 matches exist, you'll be notified to use a more specific query. If you don't have a valid tab ID, use tabs_context first to get available tabs.

Set values in form elements using element reference ID from the read_page or find tools. If you don't have a valid tab ID, use tabs_context first to get available tabs.

Extract raw text content from the page, prioritizing article content. Ideal for reading articles, blog posts, or other text-heavy pages. Returns plain text without HTML formatting. If you don't have a valid tab ID, use tabs_context first to get available tabs. Output is limited to 50000 characters by default; if it exceeds the limit it is truncated with a note giving the full size.

Execute JavaScript code in the context of the current page. The code runs in the page's context and can interact with the DOM, window object, and page variables. Returns the result of the last expression or any thrown errors. If you don't have a valid tab ID, use tabs_context first to get available tabs.

Read browser console messages (console.log, console.error, console.warn, etc.) from a specific tab. Useful for debugging JavaScript errors, viewing application logs, or understanding what's happening in the browser console. Returns console messages from the current domain only. If you don't have a valid tab ID, use tabs_context first to get available tabs. IMPORTANT: Always provide a pattern to filter messages - without a pattern, you may get too many irrelevant messages.

Read HTTP network requests (XHR, Fetch, documents, images, etc.) from a specific tab. Useful for debugging API calls, monitoring network activity, or understanding what requests a page is making. Returns all network requests made by the current page, including cross-origin requests. Requests are automatically cleared when the page navigates to a different domain. If you don't have a valid tab ID, use tabs_context first to get available tabs.

Get an accessibility tree representation of elements on the page. By default returns all elements including non-visible ones. Can optionally filter for only interactive elements, limit tree depth, or focus on a specific element. Returns a structured tree that represents how screen readers see the page content. If you don't have a valid tab ID, use tabs_context first to get available tabs. Output is limited to 50000 characters -- if exceeded, the tree is truncated at a line boundary with a note giving the full size; pass a larger max_chars, or use depth/ref_id to focus.

Resize the current browser window to specified dimensions. Useful for testing responsive designs or setting up specific screen sizes. If you don't have a valid tab ID, use tabs_context first to get available tabs.

Present a plan to the user for approval before taking actions. The user will see the domains you intend to visit and your approach. Once approved, you can proceed with actions on the approved domains without additional permission prompts.

Show a short, temporary narration ribbon in the controlled browser tab so the person watching understands the current workflow phase. Use it for meaningful phase changes, not routine clicks or keystrokes. A new narration replaces the current one.

Wait until the page is ready. By default waits for BOTH your condition and page settlement (DOM mutation rate decayed). Provide selector (CSS) or text (visible substring) with state visible|present|gone, or call with neither to wait for settlement alone. min_ms sets a minimum elapsed time; settle:false gates on the condition only. Returns elapsed_ms, settle diagnostics, and the matched element's ref for follow-up clicks. Times out with an error naming what WAS on the page.

Run a sequence of tool calls in one request. Steps execute in order; each step is validated, authorized, and audited exactly as if called individually. Step arguments may reference a prior step's structured result: $prev.field for the previous step, $N.field for step N (1-indexed), with .0-style numeric segments indexing arrays (example: $prev.results.0.ref after find). Write $$ for a literal leading $. Only tools with structured results (find, tabs_context, tabs_create, navigate, wait_for) can be referenced. Steps may not include script itself. Use wait_for between navigate and reads on dynamic pages.

Fill a form by field labels in one call. Provide fields as a map from a label, placeholder, or name attribute to the value (string, number, or boolean for checkboxes). Matching is case-insensitive and specificity-ordered; ambiguous keys are returned unmatched with candidates instead of guessed. submit:true clicks the form's own submit control after filling. Passwords are masked in the result. Falls back cleanly: anything unmatched can be filled with form_input using the refs in the result.

Resolve one visible element by ref or accessible meaning, perform one action, and return a bounded observation receipt. Use this when the target should be unique and you want to avoid a separate find, action, and wait loop. Ambiguous semantic matches are reported without acting.

Inspect or explicitly resolve the JavaScript dialog blocking one owned tab. Use status when the dialog state is unknown. Never accept, dismiss, or respond without intent from the current task.

Focus, reload, or close one tab owned by this Ghostlight session. Close is always explicit and never affects a user-owned tab or automatically deletes the containing tab group.

Upload one or multiple files to a file input element on the page. Do not click on file upload buttons or file inputs -- clicking opens a native file picker dialog that you cannot see or interact with. Instead, use read_page or find to locate the file input element, then use this tool with its ref to upload files directly.

Execute a sequence of browser tool calls in ONE round trip. Each item is {name, input} where input is exactly what you'd pass to that tool standalone. Actions execute SEQUENTIALLY (not in parallel) and stop on the first error. Use this tool extensively to quickly execute work whenever you can predict two or more steps ahead -- e.g. navigate, click a field, type, press Return, screenshot. Each tool's own permission check runs per item -- if an action navigates to a domain without permission, the next item's check fails and the batch stops. Screenshots and other images are returned interleaved with outputs; coordinates you write in THIS batch refer to the screenshot taken BEFORE this call. browser_batch cannot be nested.

Upload a previously captured screenshot to a file input or drag & drop target. Supports two approaches: (1) ref -- for targeting specific elements, especially hidden file inputs, (2) coordinate -- for drag & drop to visible locations like Google Docs. Provide either ref or coordinate, not both.

Create a short, memory-only GIF of browser work. Call start_recording, use browser tools normally, then call export; export stops capture automatically. Recording also auto-stops after 30 seconds idle or 120 seconds total. Use status to inspect state, stop_recording for an optional explicit boundary, or clear to erase immediately. Export can return the GIF to the client (download:true) or place it on the page with ref or coordinate.

Returns this server's action directory: every available action, the capability it requires (read, action, write, or execute; some require none), and a short description of what it does, plus definitions of the capability vocabulary. Use it to learn what you are allowed to do in this session. It does not read, summarize, or explain web pages.