Organize by project when one piece of work spans several apps. Give each active project a clear outcome, keep only its current windows together, and record the next action beside them. Use Mission Control or Spaces when visual layout is enough; add a project layer only when app-by-app switching no longer tells you where the work belongs.
Group by app when each application represents one kind of work. Group by project when one outcome needs windows from several apps, or the same app serves several outcomes. A small two-question test makes the choice clearer.
A useful project name identifies an outcome, includes the smallest distinguishing detail, and survives a quick recognition test. Avoid app names, broad areas, dates without context, and labels that only make sense today.
If your layout already works, add organization at the identification layer. Inventory the current windows, assign only the ones needed for each outcome, and verify that grouping did not move or resize them.
Create one active group per client delivery, include only the windows required for its next action, and keep accounts and confidential access in their proper systems. Close disposable material after handoff while preserving durable files.
Use Spaces for broad visual separation, browser profiles or accounts for identity, and project groups for cross-app outcomes. Each boundary solves a different problem, so combine only the layers you actually need.
Before switching, capture one next action and trim the current window set. On return, select the project first, verify each live window, and reopen only resources that no longer exist as windows.
Close disposable previews and duplicates, save files and links that form the record, and keep a window open only when a real next action remains. Verify unsaved work in the source app before closing anything.
When two projects use the same browser, editor, or document app, the app icon cannot identify the work. Add stable clues in titles or content, group by outcome, and verify the selected window before acting.
Start with the smallest useful set: a named outcome, a concrete next action, one working window, and at most one reference. Add a task list, project group, or separate Space only when a repeated problem justifies it.
Give the working window priority, keep only references needed for the current decision, and store durable source links outside the live layout. Close or replace references as the next action changes.
A window group records relationships among live windows. A saved session may retain resource addresses, application launch instructions, layout, or process state. Verify each layer separately before relying on a tool for restoration.
Command-Tab selects an app, Command-backtick cycles its windows, Mission Control provides a broad overview, and App Exposé narrows to one app. Use search or project groups only when those native scopes do not identify the target.
Activate the app first, then use Command-backtick to cycle its windows. If the target is minimized, full screen, on another Space, or hard to identify, inspect App Exposé, Mission Control, or the app’s Window menu.
Start with Mission Control when you recognize the window visually. Narrow to App Exposé when you know the application, or use title and project clues when thumbnails look alike. Check minimized and full-screen locations separately.
Title search is useful when you remember a document, repository, client, or webpage name but not its app or Space. Search a distinctive fragment, narrow by app, and treat duplicate or stale titles as ambiguous.
A duplicate title is insufficient identity. Add a second independent clue, verify it inside the live window, and rename the source object where possible. If no clue separates the candidates, stop rather than guessing.
Command-Tab usually returns to the previous application, while Command-backtick cycles windows in the current app. If “previous” means a specific cross-app project window, use a narrower recent or project-based list.
Build a keyboard route from project to window, keep one fallback for ambiguous titles, and avoid assigning overlapping global shortcuts. Keyboard-only navigation still requires visible focus and a safe stop when identity is unclear.
Confirm that the front app has more than one top-level window, locate the grave-accent key for the current input source, and check macOS shortcut settings. Then inspect minimized, full-screen, and tab-only cases.
Command-Tab selects an application, not a guaranteed window. Confirm the menu bar changed, then inspect the Dock, Window menu, Mission Control, and Spaces before reopening a document or creating another window.
A full-screen window occupies a Space, so app switching alone may not identify the right full-screen instance. Use Mission Control or Space navigation, verify the title, and distinguish full screen from a maximized desktop window.
List the macOS, app, and utility commands that use the shortcut, reproduce the conflict with one tool enabled, and assign one clear owner. Preserve an accessible fallback before changing a global navigation key.
Thumbnails work when windows look different; text lists work when titles, apps, clients, or files are the stronger clue. A hybrid view is useful only if both signals remain readable and the selected window can be verified.
Create only a few Spaces, place each active project’s windows together, make the desktops visually distinct, and turn off automatic rearrangement if stable order matters. Spaces organize visibility, not saved sessions.
Turn on Stage Manager, bring the working app to the center, and drag another app from the recent-app strip into the group. Remove unrelated apps and verify multi-window behavior in Stage Manager settings.
Stage Manager keeps the active app or group centered with recent sets nearby. Spaces divide work across desktops and full-screen areas. Choose one as the primary visual model, then add another layer only for a distinct need.
Turn off “Automatically rearrange Spaces based on most recent use” in Desktop & Dock settings. Verify the order after several switches and keep full-screen and multi-display behavior in the test.
macOS can switch to a Space that already contains windows for the chosen app, and Dock assignment can bind an app to a desktop. Check both before moving windows or changing shortcuts.
Use the app’s Dock menu and Assign To when most windows from that app belong on one desktop. Move an individual window through Mission Control when different windows from the same app serve different projects.
Create the destination Space first, move only the selected window, and verify its title and destination before releasing it. App assignment is a separate, broader rule.
Mission Control shows open windows on the desktop plus Spaces and full-screen areas. App Exposé narrows the view to one app. Choose the smaller candidate set that still contains the target.
Hide removes an app’s windows from view, Minimize sends one window to the Dock, and Close ends that window. The app may remain running after any of these actions, so use Quit separately when required.
Fill or maximize enlarges a normal desktop window while keeping it in the desktop context. Full screen gives the window its own Space. Choose based on whether you need desktop access and side-by-side windows.
Split View places two apps together in a full-screen Space. Desktop tiling keeps windows on a normal desktop and supports halves, quarters, and arrangements. Pick the model that matches your navigation and layout needs.
Disable edge tiling, menu-bar fill, or Option-assisted tiling individually, then repeat the same drag. Keep menu and keyboard tiling available if you only want to stop accidental pointer triggers.
Keep the action window on the primary display, references on a secondary display, and shared communication in a predictable place. Test the arrangement with separate Spaces on and off before relying on it.
Confirm the window still exists, select it through the app or Mission Control, and move or resize it onto a visible display. Reconnect the previous monitor temporarily when the window cannot be reached safely.
Use extended desktop mode, select the exact window, and drag it across the display boundary. For repeat moves, assign one conflict-free shortcut and test full-screen, minimized, and separate-Spaces cases.
When enabled, each display has its own Spaces context and the Dock is available on every display. The setting is required for Stage Manager and Split View on secondary displays, but it changes how multi-display navigation feels.
Use full screen or a half-and-half tile for the current action, switch references instead of shrinking many windows, and keep project identity in titles or groups rather than fixed screen positions.
Use the portrait display for documents, code, logs, or pages that benefit from height. Prefer stacked regions or one centered readable column, and keep wide timelines or multi-column tools on a landscape screen.
Start with a centered or two-thirds primary window, then give the remaining width to one or two references. Test readable line length, app minimum widths, and pointer travel before saving a layout.
Confirm the display arrangement and resolution first, then bring the window into view and resize it from a visible edge or use a supported move/resize command. Update saved layouts only after the new geometry is stable.
Use browser tabs for pages that travel together, a separate browser window when web material must sit beside other apps, and one project label for the cross-app working set.
Choose a tab group for one-at-a-time pages inside a single browser context. Choose a separate window when the web set must be visible, switched, or organized alongside other Mac apps.
Keep a tab for a current action, bookmark a durable reference, and save a tab group when a named set must reopen together. Move the action itself into a task.
Create a Safari profile for each browsing identity, choose the right profile before signing in, and understand which data is separate and which remains shared.
Search open tabs with a distinctive title or domain, verify the result’s window and profile, and switch to browser history only after ruling out a live tab.
Review each tab once, write a concrete action only when work remains, preserve durable references appropriately, and close pages that no longer serve the next step.
A browser workspace organizes tabs, groups, or web resources inside the browser. A desktop workspace organizes app windows at the macOS layer; neither term guarantees layout or session restoration.
Close keeps a saved group available to reopen, Ungroup leaves its tabs open but removes the group, and Delete removes the group across synced devices. Test with disposable pages first.
Detach a tab by dragging it outside the window, or move it to another same-profile Chrome window. Split only when the page needs its own system-level surface.
Use a task for one completable action, a project for a finite result needing several actions, and an area only for continuing responsibility. Skip extra hierarchy when a list is enough.
Name the action you can perform now, include the object and necessary context, and state what evidence means done. Split the work if more than one independent result remains.
Keep each task in its original project and use a Today or scheduled view to select it for the day. Reschedule unfinished work instead of creating another copy.
Honor hard constraints first, then choose the project whose next action is both meaningful and ready. Limit simultaneous work and record why anything is paused.
Stay with Reminders when its native task structure and alerts cover the project. Add another tool only for a verified missing capability such as cross-app window return.
Create, rename, schedule, and complete each commitment in one task system. Let the second tool organize current windows without copying the task database.
Inspect each active project, restore one executable next action, verify important resources can still be reached, and pause or close work that no longer deserves attention.
Create a minimal working set for one code change, repeat the repository and environment clue across windows, and close obsolete logs or preview sessions after verification.
A workspace is the folder or saved set of folders and workspace settings loaded in VS Code. A window is one running UI instance that hosts a workspace.
Use multi-root for related folders that share one editing and search context. Use separate windows when projects, environments, settings, or screen placement should remain independent.
Use tabs for sequential shells in one context, panes for commands that must be watched together, and windows for independent projects, displays, or environments.
Show a short project and environment label in the terminal title or prompt, keep the working directory visible, and update the clue when the shell changes purpose.
Read the change context first, review the diff in focused order, reproduce relevant behavior locally, and keep feedback and validation evidence beside the code.
Record one reproducible symptom, align logs and code around a testable hypothesis, preserve timestamps and inputs, and reset the workspace when the hypothesis changes.
Repeat the environment identity in browser profiles, terminal titles or prompts, and editor workspace names, then verify credentials at the action boundary.
Separate source discovery from close reading, capture claim-level notes with links, write from the notes, and keep only the sources needed for the current section open.
Choose a single feedback source, label the current design version, pair each comment with its evidence and preview, and archive obsolete references after approval.
Review the notes for decisions, commitments, owners, and dates; create only actionable tasks; link durable material; and close the meeting workspace after handoff.
Keep the lesson visible with one note destination and one practice environment, record time or section checkpoints, and close unrelated course tabs after each session.
Give inputs and generated outputs separate locations, label Finder windows or folders clearly, compare the expected files, and move approved results only after verification.
Share one tab or window when possible, verify the exact source and permissions, move private notifications and unrelated windows away, and test what attendees will see.
Use macOS tiling when its built-in arrangements cover the task. Add Rectangle only when verified extra shortcuts or positions justify another permissioned utility.
Moom saves window positions for visible windows and can apply layouts to matched apps or recent windows, with launch support subject to limitations. Verify contents before saving.
Use Command-Tab for simple app switching. Consider AltTab when direct window selection and previews solve a repeated same-app problem and its permissions are acceptable.
Choose rcmd for rapid known-target switching and search; choose project grouping when the repeated problem is returning to several current windows for one goal.
Use manual placement for changing, exception-heavy work. Use automatic tiling when a repeatable rule consistently handles the active window set and you accept reflow.
Choose a launcher when the project starts from files, URLs, and apps. Choose a window organizer when the resources are already open and the problem is selecting the correct live windows.
Use native Chrome groups for lightweight tab sets. Evaluate Workona when verified workspace resources and project workflows justify another account and browser layer.
Use dimming when the current window is clear but surrounding contrast is distracting. Use project grouping when the challenge is finding and returning to the right cross-app set.
Screen Recording permits access to on-screen content. Whether a tool captures previews, saves them, or uploads them depends on that product’s current implementation and privacy policy.
Accessibility can let a trusted app identify and focus windows. Screen Recording covers visual previews. Grant only the access needed and test the feature after approval.
First check whether the window entry exists. Then verify the correct app has current Screen Recording access, refresh the app, inspect filters and window state, and test a normal window.
Record the selected row and the focused window, then test duplicate titles, changing sort/filter rules, and whether the original window was closed and recreated with a new identity.
Ask three independent questions: what works without a network, whether an app identity is required, and where each data type is stored. Verify first-use and recovery separately.