Pilot

Navigate to a URL. Returns HTTP status code and final URL.

Go back in browser history.

Go forward in browser history.

Reload the current page.

Get accessibility tree snapshot with @eN refs for element selection. Use refs in click/fill/hover/etc. Use include_cursor_interactive to find non-ARIA clickable elements (@cN refs).

Compare current page state against the last snapshot. Returns unified diff showing what changed. First call stores baseline.

Take a screenshot with red overlay boxes and ref labels at each @eN/@cN position. Requires a prior pilot_snapshot call.

Click an element by @ref (from snapshot) or CSS selector. Auto-routes clicks to selectOption.

Hover over an element by @ref or CSS selector.

Clear and fill an input/textarea by @ref or CSS selector.

Select a dropdown option by value, label, or visible text.

Type text into the currently focused element (character by character).

Press a keyboard key (Enter, Tab, Escape, ArrowDown, Backspace, etc.).

Drag from one element to another.

Scroll element into view, or scroll to page bottom if no ref provided.

Wait for element visibility, network idle, or page load.

Upload file(s) to a file input element.

Extract clean text from the page (strips script/style/noscript/svg).

Get innerHTML of a selector/ref, or full page HTML if none provided.

Get all links on the page as text + href pairs.

Get all form fields on the page as structured JSON.

Get all attributes of an element as JSON.

Get computed CSS property value for an element.

Check element state: visible, hidden, enabled, disabled, checked, editable, focused.

Get console messages from the circular buffer.

Get network requests from the circular buffer.

Get captured dialog (alert/confirm/prompt) messages.

Run a JavaScript expression on the page and return the result. Supports await.

Get all cookies as JSON.

Get localStorage + sessionStorage as JSON (sensitive values redacted). Optionally set a localStorage key.

Get page load performance timings (DNS, TCP, TTFB, DOM parse, load).

Take a screenshot of the page or a specific element.

Save the current page as a PDF.

Take screenshots at mobile (375x812), tablet (768x1024), and desktop (1280x720).

Text diff between two URLs — compare staging vs production, etc.

List all open browser tabs with URLs, titles, and active status.

Open a new browser tab, optionally navigating to a URL.

Close a browser tab by ID (or current tab if none specified).

Switch to a specific browser tab by ID.

Set the browser viewport size.

Set a cookie on the current page domain.

Import cookies from a real Chromium browser (Chrome, Arc, Brave, Edge, Comet). Decrypts from browser cookie database and adds to the headless browser session.

Set a custom request header. Sensitive values are auto-redacted in the response.

Set the browser user agent string. Recreates the browser context, preserving cookies and state.

Configure how dialogs (alert/confirm/prompt) are handled.

Open a visible (headed) Chrome window with all current state — cookies, tabs, localStorage. Use when headless mode is blocked by CAPTCHAs, bot detection, or complex auth. The user can solve it manually, then call pilot_resume to continue.

Resume control after user handoff. Takes a fresh snapshot of the current page state.

Close the browser and clean up all resources.

List all frames (iframes) on the page. Use pilot_frame_select to switch context into an iframe for snapshot/interaction.

Switch context to an iframe by index or name. After switching, pilot_snapshot/click/fill will operate inside that iframe. Use pilot_frames to list available frames.

Switch back to the main frame. Use after interacting with an iframe.