Preferred NexBrowser OpenAPI tool for requests such as 'use NexBrowser to show or count my windows'. Lists managed NexBrowser environments/profiles. Never reports Chrome, Edge, the Codex in-app browser, operating-system application windows, or browser tabs.

List proxy resources available in the active NexBrowser workspace so a proxyId can be selected for window creation or binding. Credentials are never returned.

Import custom proxy resources from Desktop-compatible text lines the user supplied in the current request. Supported protocols are HTTP, HTTPS, and SOCKS5. Credentials are never returned.

Create one custom proxy resource from protocol, host, and port, or pass items to create several. Prefer nex_proxy_import when the user pasted Desktop-compatible lines. Credentials are never returned.

Create multiple custom proxy resources from an items array. Each item needs host and port. Credentials are never returned.

Update one existing custom proxy resource. Pass only the fields that must change. Credentials are never returned.

Delete one or more custom proxy resources. Bought channel proxies may be rejected by Desktop. Credentials are never returned.

Probe a proxy resource or an unsaved host:port for exit IP, location, and timezone. Credentials are never returned.

Create NexBrowser environments in the desktop app's active workspace. Every unset field falls back to the desktop create-window defaults (fingerprint, preferences, layout, startup page), and each created window gets its own generated fingerprint, so pass only the fields that must differ.

Bind one proxy resource to one or more closed NexBrowser windows. Pass proxyId=0 to remove the binding. Running windows are rejected and must be closed first.

Bind one or more workspace catalog platform accounts to one or more closed NexBrowser windows. This replaces the existing window binding. Pass accountIds=[] to remove it. Running windows are rejected and must be closed first. After opening, use nex_browser_accounts to fill.

List the window groups of the active NexBrowser workspace with the window count of each, so a groupId can be selected for window creation, moving, or deletion. The first row is always the ungrouped bucket with groupId 0.

Create one window group in the active NexBrowser workspace. Group names must be unique inside the team; a duplicate name is rejected. Use nex_browser_group_modify to rename it, or nex_browser_move_to_group to put windows into the new group.

Rename or reorder one custom window group in the active NexBrowser workspace. Group names must stay unique inside the team. The ungrouped bucket (groupId 0) cannot be changed.

Delete one custom window group from the active NexBrowser workspace. The windows inside it are not deleted; they become ungrouped. The ungrouped bucket (groupId 0) cannot be deleted.

Move one or more NexBrowser windows into a window group. Pass groupId=0 to move them out of any group. Groups are window metadata, so running windows can be moved without closing them first.

Start NexBrowser environments and return per-window status. For multi-window tasks, pass every window ID in one windowId array so Desktop can batch-start and tile them before creating automation sessions.

Inspect the status of running NexBrowser environments. Omit windowId to list every running window.

Close one or more NexBrowser environments.

List autofill-capable credentials in the vault plugin of open NexBrowser environments, including bound platform accounts and user-saved passwords. Usernames are masked and secrets are never returned; call nex_browser_fill_account with an accountId from this result.

List platform accounts in the workspace catalog so an accountId can be selected for window creation or later vault fill. This is not the open-window vault; use nex_browser_accounts after a window is running. Secrets are never returned.

Create one platform account in the workspace catalog, or pass items to create several. Pass platformUrl; username, password, and 2FA secret are optional and write-only. This does not fill a login form.

Create multiple platform accounts from an items array. Each item needs platformUrl. Secrets are write-only and never returned.

Update one platform account in the workspace catalog. Pass only the fields that must change. Password and 2FA secret are write-only and never returned.

Delete one or more platform accounts from the workspace catalog. This does not close windows or remove vault copies already stored in an open window.

Connect to one NexBrowser window using the active team by default and create an isolated automation session. For multiple windows, first call nex_browser_open once with all window IDs, then connect each window with startIfNeeded=false.

Disconnect an automation session without closing the NexBrowser environment.

List tabs and the active stable page ID for an automation session.

Open a new tab and optionally navigate it.

Select a tab by stable page ID.

Close a tab by stable page ID.

Navigate the selected page and return its refreshed accessibility snapshot.

Page reloaded.

Navigated back.

Navigated forward.

Click or double-click an element from the latest snapshot or a CSS selector.

Hover over an element.

Type text into an element using sequential key events.

Fill one or more form controls in order.

Ask the NexBrowser vault plugin to autofill the current login page with a stored credential. Call nex_browser_accounts first and pass only an accountId returned for this open window; the plugin chooses the visible fields and keeps passwords and 2FA secrets outside MCP. Call again after each navigation in a multi-step sign-in. This tool never submits the form.

Fill literal login credentials that the user explicitly supplied in the current request into visible fields. Prefer nex_browser_fill_account when a stored vault credential is available, because literal values cross the model and MCP request boundary. Never store or repeat these values. This tool never submits the form.

Select one or more values in a select control.

Check a checkbox or radio control.

Uncheck a checkbox or radio control.

Press a key on the selected page or focused element.

Hold a keyboard key down until nex_browser_key_up is called.

Release a keyboard key held by nex_browser_key_down.

Drag one snapshot ref or selector to another.

Scroll the page, optionally bringing a target into view first.

Wait for a target, text, or a bounded duration.

Resize the active page viewport.

Capture an accessibility/ARIA snapshot with short-lived actionable refs.

Page title.

Page URL.

Read visible text from the page or one target.

Read console messages buffered by this automation session.

Read network requests buffered by this automation session.

Read one buffered network request by its 1-based index.

List pending dialogs or accept/dismiss one dialog.

Evaluate JavaScript when the dedicated evaluate permission is enabled.

Run page JavaScript when the separate runCode permission is enabled.

Capture a screenshot and return MCP image content when it is small enough.

Upload approved local files through a file input target.

Drop approved files or MIME-typed text data onto a target.