Browser Connector

​

Overview

MuleRun Browser Connector is a Chrome extension (displayed as MuleRun Chat) that connects your local Chrome browser to the MuleRun platform. Once installed and signed in, Super Agent can open new tabs in your local browser to carry out tasks—using your already-logged-in accounts to visit sites, read page content, fill forms, click buttons, and scrape data. The Agent operates the browser through the extension, and your credentials and local sessions never leave your device. Browser Connector has two entry points:

• Chrome extension (MuleRun Chat) — Once installed, a Mule icon appears in the browser toolbar. Clicking the icon, using the right-click menu, or pressing the shortcut (⌘U on Mac / Ctrl+U on Windows) opens MuleRun’s sidebar chat on the right of the page, with the current page brought in as context. The extension’s Options page manages the local browser name, enabled state, auto-approve, appearance, and UI language.
• Connector on mulerun.com — In the web chat input’s connector panel, toggle on My Browser to route the session through your local browser. Settings → Browsers lists all browser extensions currently connected under your account and lets you enable/disable each one. If the extension isn’t installed, opening /my-browser shows the install guide and download package.

​

Target Users

​

Use Cases

• Start a chat while browsing — Press ⌘U / Ctrl+U or click the toolbar icon to open the sidebar and send an instruction; the Agent automatically brings in context like the current tab’s URL and selected text.
• Ask about selected text in place — Highlight text on a page, right-click “Chat about this with MuleRun,” and the sidebar opens with that quote attached.
• Scrape content that requires login — Install the extension and toggle on My Browser in chat to let the Agent open sites with your signed-in state (e.g., subscribed channels, DMs, order history).
• Auto-fill web forms — Have the Agent open a page and fill fields from your draft, then you confirm whether to submit.
• Multi-account isolation — Sign in to different MuleRun accounts in different Chrome profiles; each account only sees and uses its own browsers—they’re invisible to each other.
• Intranet / SSO systems — Use your machine’s existing single sign-on session to access internal company systems.
• Supervise long-running tasks — Tabs the Agent operates are grouped under a tab group named “MuleRun Chat” with a status icon in the group title; click Takeover at the bottom of the tab to take control anytime.
• Same account across multiple computers — Install the extension on home and office machines under the same MuleRun account; use Browsers to control which are available—if several are online, the Agent will ask which one to use.

​

Quick Start

The shortest path: install the extension and bring up the sidebar chat from a regular web page.   1. In Chrome, open MuleRun’s tutorial page (/my-browser) and click Download Extension to get the install package.   2. Unzip the ZIP into a local folder. Visit chrome://extensions, enable Developer mode, click Load unpacked, and select the folder.   3. Pin the Mule icon in the Chrome toolbar. Click it to open the sidebar and complete MuleRun sign-in as prompted.   4. Open a regular web page (e.g., an article or product detail page) and press ⌘U (Mac) / Ctrl+U (Windows), or click the toolbar Mule icon.   5. The sidebar chat opens on the right of the browser, with a context card for the current page above the input.   6. Type an instruction (e.g., “Summarize the key points of this page”) and send—the Agent starts working in your local browser. The sidebar shows the Agent’s replies and progress. If the shortcut does nothing, see the FAQ below.

​

Detailed Guide

1. Installation and setup Install the extension. Open /my-browser in Chrome → click Download Extension → unzip the ZIP → visit chrome://extensions, enable Developer mode, click Load unpacked, select the folder → pin the Mule icon → click it and sign in. Constraints: only Chrome on desktop is supported (Edge/Opera/Firefox/Safari and mobile are not); Chrome shows a “Disable developer-mode extensions” prompt at startup—click Cancel to keep it; the extension and mulerun.com must use the same Chrome profile, or login won’t be recognized. Configure browser settings. Open the options page (right-click the Mule icon → Options):

• Browser Name — auto-generated two-word name; click the dice to regenerate.
• Enable this browser — when off, no tasks are routed here. (Free accounts: 1 enabled at a time—enabling a second shows “Disable another browser first.”)
• Auto-approve authorization — skip the confirmation card in mulerun.com chat.
• Appearance — Light / Dark / System (affects sidebar chat and options page).
• Language — switch UI language.

2. Starting a chat From a web page (sidebar). Four entry points open the sidebar with the current page as context:

• Click the Mule icon — toggles open/closed.
• ⌘U / Ctrl+U — opens with current tab.
• Right-click a blank area → “Chat about this page with MuleRun.”
• Select text → right-click “Chat about this with MuleRun.”

A context card above the input shows the URL, title, or selected quote; the Agent can read this tab within the session without further authorization. Click Refresh page context to switch context after changing tabs. If a session is already ongoing, triggering again prompts Join current chat or Start a new chat (stopped/completed/errored sessions go straight into a new one). From mulerun.com chat. Click the connector icon (cable shape) below the chat input—My Browser sits at the top. If not installed, an Install button opens a guided dialog (with the chrome://extensions/ link, download, and 5-step tutorial); if installed and enabled, the toggle routes browser tasks to your local extension. (The toggle can’t be switched while a task is running—stop it first.) 3. Authorization One-time task authorization (Authorize). The first time the Agent uses your local browser in a mulerun.com chat session, a card appears with a 120-second countdown:

• Authorize — runs the task immediately; further browser calls in this session won’t ask again.
• Reject — falls back to the cloud browser; re-authorizing requires a new chat session.
• Free users see an extra “Free plan allows only one browser at a time” hint with a Manage browsers link.

Authorization is per chat session, valid for 30 days. (Sessions started from the extension sidebar skip this card—a green Always on badge confirms it.) Security note: the extension can access all browser pages—avoid handing it sensitive pages like online banking or internal systems during authorized tasks. Auto-approve. Turn on Auto-approve authorization in options to skip the card for sessions routed to this browser. (A green rocket icon marks auto-approve in options and on Browsers; each browser’s toggle is independent.) 4. While a task is running Watching the Agent work. Operated tabs are grouped under a tab group named “MuleRun Chat,” with a status icon in the group title. The active tab shows a semi-transparent overlay and a virtual cursor that follows the Agent’s actions, with a Takeover button at the bottom center. Takeover. Click the Takeover button to stop the entire session immediately and return control. (Hard stop—can’t be resumed; start a new task to continue. Just closing the Agent’s tab only interrupts that tab, and the task may fail or need restarting.) 5. Managing multiple browsers Browsers list. Each Chrome profile with the extension registers as a separate browser. Open Settings → Browsers to see all connected extensions—each row shows:

• Status dot — green = enabled, gray = disabled.
• Browser name — the one labeled Current is the browser you’re viewing from.
• Connection info — duration, ongoing sessions, auto-approve marker.
• Enable toggle — turn off to stop receiving tasks. (Hitting the tier limit triggers a Manage browsers prompt to disable one first.)

Routing rule. When multiple enabled browsers are online, the Agent asks which to use; it auto-routes only when exactly one is online. 6. Stopping use and uninstalling

• Temporarily disable — turn off Enable this browser in options or the row toggle in Browsers (extension stays signed in; re-enable to resume).
• Uninstall — remove from chrome://extensions; the browser disappears from Browsers (which only shows online extensions). Reinstalling registers as a new connection, not the original record.
• Revoke a specific authorization — no UI for this; authorizations expire automatically after 30 days.

​

Plans, Quotas, and Limits

Browser Connector is available on all tiers (Free / Plus / Super / Pro / Team Plus). Auto-approve is available on all tiers. One-time task authorization is kept for 30 days uniformly, then expires automatically. Browser Connector task usage is billed under MuleRun’s standard Agent task rules.

​

FAQs

Q: Pressing ⌘U / Ctrl+U does nothing—why? A: Most often the shortcut is taken by another Chrome extension or by the system. Go to chrome://extensions/shortcuts, find MuleRun Chat’s “Open MuleRun chat for the current page,” and assign a non-conflicting shortcut. The other case: the current page is a system or local page (chrome://, chrome-extension://, devtools://, edge://, about:, file://)—Chrome doesn’t let extensions act on these, so the shortcut won’t bring in page context; it will only open or restore the sidebar chat. Q: I installed the extension but the sidebar keeps saying “Sign in required”—why? A: Usually the Chrome profile running the extension isn’t signed in to MuleRun, or the signed-in account is in a different Chrome profile. Click Sign in in the sidebar and complete login with your MuleRun account. Q: When the Agent is working, will it read content from other tabs I have open? A: It depends on how the session was started. When started from the extension shortcut, right-click menu, or toolbar icon, what you’re handing to the Agent is the current tab—it reads and operates in that tab directly. When started from the mulerun.com web chat, the Agent opens new task tabs itself by default and doesn’t read other open tabs. (The extension has permission to access all browser pages—see the security note under One-time task authorization for sensitive pages.)