MicsApp Web Terminal: a browser terminal and server file workspace guide

MicsApp Web Terminal provides a real server shell in a browser, with persistent sessions, tabs, split panes, a file browser/editor, saved commands, notes, voice input and optional remote desktops. Your administrator supplies the server URL and account. The project repository contains deployment documentation; this guide explains the everyday interface and the companion mics_cli client.

Sign in and run your first command

1. Open the HTTPS address your administrator provided, for example https://terminal.example.com.
2. Enter Username and Password, then Sign in.
3. Wait for the terminal and prompt. Click the terminal pane to focus it.
4. Enter pwd and press Enter to see the current server directory. Enter ls to list files.
5. Use + New Tab for another shell without closing the first.

The account is a server system account, not a new website registration. Login errors can mean an incorrect password or an account that the administrator has not enabled. The shell uses the account's permissions; a browser connection does not grant access to other users' files.

All screenshots show an isolated sandbox with a made-up demo user, fake remote-server entries and a small weather-station project. They do not show the owner's server files or credentials. Local terminal, sessions, commands, notes and file editing were exercised; external SSH, physical-device dictation and VNC depend on services outside this walkthrough.

Understand the toolbar

Buttons that do not fit on a narrow display move into Menu. The terminal must have focus for ordinary typing and terminal shortcuts. Clicking a text field, dialog or another pane changes where input goes.

Open, rename, restore and close tabs

1. Choose + New Tab to create another server shell.
2. Select a tab name to bring it forward.
3. Double-click its name to rename it. The server session name updates for other devices attached to it.
4. Use an app tab's close control only when you intend to end its underlying shell and programs.
5. Open Sessions to attach a shell that is still running on the server.

Closing the browser page preserves tmux shells. Closing an app tab with its × explicitly kills the associated tmux window, so its programs end on every attached device. Sessions → ✕ and entering exit also end a shell. A server restart can end processes unless the server has separate restart arrangements.

The browser remembers its open app tabs for reload, while Sessions lists actual server sessions. If an old tab is blank, try Reconnect or attach the session through Sessions rather than immediately starting a replacement job.

Arrange split panes and tiles

1. Choose Split Right for side-by-side panes or Split Down for a vertical stack.
2. Click a pane to make it active, then choose the tab to display in it.
3. Drag a divider to resize the panes.
4. Choose Unsplit to return to one pane.

For several existing sessions, open Sessions and choose Tile all, or tick specific sessions and choose Tile selected. This arranges sessions into a grid. Splitting is disabled below 768 pixels wide, so use tabs on a phone or increase the available width on a tablet.

Keep the active pane clear before typing: a command runs in the focused shell, even when another pane looks more prominent.

Attach sessions from another device

Open Sessions to see live shells for your account, with names, slots, running command, activity and pane count. Open attaches one to this browser; Focus selects a session already open here. open here identifies local attachments and remote identifies remote SSH tabs. Use refresh to update the list.

Start a harmless long-running command on one device, then open that same account/session from another device to watch it continue. Both attachments are live views of the same shell, so typing in one can affect the other.

Use the row's ✕ only when you intend to end the underlying session and its programs for every device. The app-tab close control also kills its associated shell; closing the browser page alone preserves it. The guide's screenshots show demonstration sessions only.

Send a command to one or several sessions

The bottom of the Sessions dialog includes Command to send…, a quick-command picker and a Notes button.

1. Type a command, or choose the lightning-bolt picker to insert a saved command.
2. Review it before sending.
3. With no rows checked, Send to active targets the focused terminal.
4. Tick rows to change the default to Send to selected (n).
5. Send to all targets all eligible local sessions after confirmation.

Sending includes Enter, so the command executes rather than merely appearing as editable text. It can run in a server session that is not an open browser tab. Remote SSH and graphical desktop tabs are skipped by the local-session broadcast path.

The checked selection determines the target; clear it to return to the active shell. Start with a simple pwd or printf when learning the composer so you can confirm the destination.

Write, save and dictate Notes

1. In Sessions, choose the Notes button beside the command field.
2. Enter an optional Note title and your text. The editor supports multiple lines without phone autocorrection, capitalization or spellchecking changing commands.
3. Choose Save note; when editing an existing note the control can read Save changes.
4. Select a saved note to reopen it; New starts another and Delete removes the selected note.
5. Use Send to active, Send to selected (n) or Send to all when the text is ready to execute.

Editing does not send anything to a terminal. Sending saves the latest edits first. The note editor allows up to 32,768 characters and its title up to 200. Notes are stored for the account on the server, making them available on another logged-in device.

Voice starts dictation; Stop ends it. Desktop uses browser speech recognition when available. Mobile captures phrases for server transcription and removes temporary audio afterward. Microphone permission, HTTPS and the configured transcription service are required. Dictation produces text to review; shell syntax, punctuation and names still need checking before Send. Enter uses the default target; Ctrl+Enter/Cmd+Enter sends to all from the composer.

Browse folders, paths and recent locations

Choose Files to open the sidebar, or use Ctrl+Shift+E.

The file browser acts on the server filesystem under your account's permissions. It is not a separate cloud drive or a local-computer file manager. A path that works in one server's file browser may not exist on a remote connection.

Upload, create, download, rename and delete files

1. Navigate to the intended destination folder.
2. Choose Upload and select files, or drag files onto the sidebar.
3. Choose + Folder and enter a name for a new directory.
4. Use Paste Img to save an image from the clipboard after supplying a filename. Browser clipboard permission and supported image content are required.
5. Use a row's Download, Rename or Delete action. Delete asks for confirmation.

Uploads and filename changes affect the selected server directory. Check destination and overwrite prompts before accepting them. Download also appears in the preview toolbar. If browser or proxy upload limits reject a large file, split it or use an administrator-approved transfer route.

Preview and edit files in the browser

Select a file to open a preview window. Several previews can be open, dragged by their title bars and enlarged with the preview's fullscreen control.

For text editing, choose Edit, change the content, then Save to write it back to the actual file. Cancel leaves the file unchanged. If you only wanted a local copy, choose Download instead.

A format that the browser cannot decode may still download correctly. Some documents require external resources or scripts that do not work inside a preview; use the original application when necessary.

Build and search your quick-command library

Open Commands to see saved commands with names, text and tags.

1. Type in Search commands by name, command, or tag..., or select a tag chip.
2. Select a command to submit it to the active terminal.
3. Choose + Add, enter Name, Command and comma-separated Tags, then Save.
4. Use a row's edit/delete controls to maintain an existing entry.

Selecting a command from the main library runs it. The Sessions lightning-bolt picker instead helps compose a command for a chosen target, giving you an opportunity to review it before Send.

Import, export and synchronize commands

Export downloads JSON. Import selects a compatible JSON array with at least name and command per entry, plus optional tags. In the browser, matching names are merged case-insensitively: imported text/tags update the corresponding command, while new names are added. Review imported commands before running them.

Sync merges your library with the shared repository configured by the administrator. It adds/updates entries rather than propagating deletions, so a command deleted locally can return if it still exists in the shared repository. An unconfigured/unreachable repository produces a sync error; local saved commands still work.

Set fonts, cursor, scrollback and app name

Open Settings, adjust the fields, then Apply to All Tabs. Applying reconnects terminal frames with the chosen appearance while preserving server sessions.

For example, [{count}] {tab} — {user} includes tab count, active tab and username. Terminal appearance and title settings are saved in this browser, so a second device can use a different presentation. The installed app name is also persisted through the account service for the manifest.

Choose a theme or custom colors

Open Themes and choose Default Dark, Light, Monokai, Solarized Dark, Dracula, Nord, Gruvbox or Tokyo Night. Alternatively set Background, Foreground, Cursor and Selection colors. Choose Apply to All Tabs to apply and remember the settings.

Theme changes affect terminal appearance; programs running inside the terminal may also have their own color configuration. If a shell application sets its own background or styling, inspect that application's preferences too.

Open configured remote servers and desktops

Open ▾ beside + New Tab. Administrators publish remote-server entries and their status; refresh the list if a new server is missing.

First-time SSH can ask for a host-key confirmation, then a password if no key is available. The app does not make an unavailable server reachable or supply a destination account. A blocked embedded Web Terminal may need the destination's framing configuration corrected by its administrator.

Desktop opens an optional noVNC graphical desktop in an app tab. Its side controls handle fullscreen and desktop-specific settings. A working server display and VNC service are required; the shell/file browser do not depend on VNC. The walkthrough did not connect to a real remote desktop.

Work from phones and tablets

The compact toolbar retains essentials and moves the rest under Menu. The special-key strip provides keys such as Esc, Tab, Ctrl, Alt, arrows and editing/navigation keys that a phone keyboard may omit. Modifier toggles apply to the next relevant key rather than changing the server's account or shell.

Long-press terminal text for selection/copy; the copy dialog provides selectable text and a Copy button when direct clipboard interaction is unreliable. Notes is useful for pasting long commands, multi-line text or reviewed voice dictation. Mobile browser keyboard and clipboard behavior varies; physical iOS/Android acceptance was not part of the emulation checks.

Use the browser's Add to Home Screen/install action where available. A network connection is required for the remote shell even when the interface is installed; installation does not make the server available offline.

Use desktop keyboard shortcuts

Browser and OS reservations can intercept these; toolbar controls provide the same actions.

Create and revoke API tokens

1. Open Settings → API Tokens → + New Token.
2. Give it a descriptive name, such as laptop.
3. Copy the value immediately; the interface shows the secret only once.
4. Supply it as Authorization: Bearer <your-token> to the HTTP API.
5. Choose Revoke beside an old token when it should stop working.

Tokens permit programmatic file, command and account-authorized API actions. They cannot open the browser terminal's /ut/ WebSocket directly. The CLI's interactive shell uses the separate authenticated /api/shell path. Treat a token as access to the account; the screenshots intentionally contain no token value.

Use mics_cli and multiple server profiles

Obtain the standalone client binary or the cli/ package from your administrator. The JavaScript package requires Node 18+; a packaged binary includes its runtime. Authenticate and check the target before issuing commands:

mics_cli login --url https://terminal.example.com
mics_cli whoami
mics_cli exec "pwd"
mics_cli ls ~

mics_cli exec "wc -l" --stdin-file ./readings.csv
mics_cli exec "ls" --cwd /tmp --timeout 60
mics_cli upload ./report.txt ~/uploads
mics_cli download ~/notes.txt -o ./notes.txt
mics_cli quick-commands export -o commands.json
mics_cli quick-commands import commands.json
mics_cli --profile demo exec "uname -a"

exec waits for completion: its default timeout is 30 seconds, with a 300-second maximum. Use an interactive shell or a deliberately detached server job for longer work. exec's exit code is the remote command's exit code, which makes it suitable for scripts. The HTTP /api/exec operation similarly accepts command, working directory, timeout and optional stdin.

Each login can create a saved server profile. use changes the active profile; --profile NAME targets one invocation. An explicit token or a local/parent .env can override the profile, so use whoami to check the URL and token source. logout forgets saved local authentication; revoke a token to invalidate it on the server.

Read Help and report a problem

Help opens the illustrated user manual served by your installed server, so it matches that version. About shows app version, build and username and offers View Manual. The repository manual and CLI reference provide additional examples when repository access is available.

No rendered promo/tutorial video or native store listing was found in the reviewed repository. The main supporting material is the running interface and illustrated manual.

Troubleshooting

When reporting an issue, include About's build, device/browser, exact action and error. Keep passwords, token values and private terminal output out of screenshots sent to a public channel.