GUI install, update & removal
Install, update, license and cleanly remove apps, add-ons and extensions via stores, DMGs and wizards.
20 project-authored proposals · taxonomy v1.1 · 0 new benchmark runs. Verifiers and partial-credit rules below are designs, not executed results.
- Typical applications
- App Store · Microsoft Store · installers
- Class GUI-need label
G2· GUI-essential
Task proposals · 20
Expand to inspect the complete original wording and proposed checks. Original proposal text remains English in both interfaces.
R01.03.01.01-T01 Install an app from a DMG and clean up
Original instruction
Install Foo from ~/Downloads/Foo-2.4.dmg by dragging it into /Applications, eject the disk image, move the DMG to the Trash, and launch Foo once to confirm it opens.
- Platforms
mac· macOS- Applications
- Finder · installers
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
drag· Drag & dropclick· Point & click- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
/Applications/Foo.app exists, CFBundleShortVersionString 2.4, codesign --verify passes; no Foo volume in `mount`; DMG in ~/.Trash; Foo process seen in launch log. Terminal withheld; eslogger detects cp/ditto/hdiutil -> 0.
Proposed partial credit
50% app installed correctly; 20% image ejected; 15% DMG trashed; 15% launched
R01.03.01.01-T02 Custom installer options without bundleware
Original instruction
Run C:\Fixtures\NoteKit-Setup.exe and install for the current user only into D:\Apps\NoteKit, decline the bundled 'SearchBar' offer, skip the desktop shortcut, and keep the Start menu shortcut.
- Platforms
win· Windows- Applications
- installers
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G3· GUI-only- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
HKCU Uninstall\NoteKit InstallLocation=D:\Apps\NoteKit (no HKLM entry); no SearchBar files/uninstall key; no Desktop .lnk; Start menu .lnk present. cmd/PowerShell blocked by AppLocker; Sysmon detecting silent-install switches (/S, /D=) in the command line -> 0.
Proposed partial credit
30% per-user install location; 35% bundleware declined; 35% shortcut choices
R01.03.01.01-T03 Install two apps from the Microsoft Store
Original instruction
Using the Microsoft Store app signed in with the test account, install Microsoft PowerToys and Python 3.12. Do not install any other app or edition (e.g. not Python 3.13).
- Platforms
win· Windows- Applications
- Microsoft Store
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
type· Text & formsclick· Point & click- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
Get-AppxPackage/winget list (grader) shows PowerToys (msstore source) and PythonSoftwareFoundation.Python.3.12; no other new packages vs seed. winget/PowerShell blocked by AppLocker; Sysmon detects winget install by the agent -> 0.
Proposed partial credit
45% PowerToys; 45% correct Python edition; 10% no extra installs
R01.03.01.01-T04 Selective App Store updates with auto-update off
Original instruction
In App Store settings turn off automatic updates, then from the Updates list install only the pending Keynote update and leave the Numbers and Xcode updates pending.
- Platforms
mac· macOS- Applications
- App Store
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
licensed· Licensed pro app- Proposed difficulty
- 3 / 5
Proposed verifier
com.apple.commerce AutoUpdate=0 (and SoftwareUpdate app auto-update off); Keynote version equals latest test catalog; Numbers and Xcode versions unchanged. Terminal withheld; eslogger detects mas/softwareupdate CLI -> 0.
Proposed partial credit
30% auto-update off; 50% Keynote updated; 20% others unchanged
R01.03.01.01-T05 Completely remove a sync app with its helpers
Original instruction
Remove SyncHelper entirely: quit it, use its bundled Uninstall.app (admin password in fixtures/admin.txt), then remove leftovers its uninstaller misses: login item, LaunchAgent, privileged helper, Application Support folder and preferences. Keep ~/Documents/SyncHelper Exports.
- Platforms
mac· macOS- Applications
- Finder · installers · System Settings
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicknav· Pan/zoom/scrolltype· Text & forms- Work product
WP08· App/system state- Conditions
risk· Irreversible/risky- Proposed difficulty
- 5 / 5
Proposed verifier
Absent: /Applications/SyncHelper.app, ~/Library/LaunchAgents/com.synchelper.*, /Library/PrivilegedHelperTools/com.synchelper.helper, App Support, prefs plist, BTM entry; launchctl list has no label; Exports folder hash intact. Terminal withheld; eslogger detects rm/launchctl/sfltool exec -> 0.
Proposed partial credit
20% app and process gone; 50% five leftover locations clean (equal share per location); 30% user exports preserved
R01.03.01.01-T06 Uninstall an app whose uninstaller is blocked
Original instruction
Uninstall 'LegacyCAM 4' from Settings > Apps. Its uninstaller fails while LegacyCAM's tray process and Windows service are running; stop them properly, finish the uninstall, and choose to keep the user configuration folder when asked.
- Platforms
win· Windows- Applications
- Windows Settings · Task Manager · Services
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Uninstall key gone; Program Files\LegacyCAM removed; service LegacyCAMSvc deleted; %APPDATA%\LegacyCAM\config.json kept (hash); no LegacyCAM processes. cmd/PowerShell blocked by AppLocker; Sysmon detects msiexec /x /qn or sc delete by the agent -> 0.
Proposed partial credit
20% processes/service stopped; 50% uninstall complete; 30% user config kept
R01.03.01.01-T07 Load an unpacked Chrome extension
Original instruction
Install the unpacked extension in ~/Desktop/ext-tabtimer into Chrome via chrome://extensions with Developer mode, pin it to the toolbar, and allow it in Incognito.
- Platforms
mac· macOSwin· Windowsweb· Any OS (web)- Applications
- Chrome
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Chrome Preferences extensions.settings has the extension path with state enabled, location unpacked, incognito=true; extensions.pinned_extensions contains its id. Terminal/PowerShell withheld; exec audit (eslogger/Sysmon) detecting Chrome launched with --load-extension or Preferences edited outside Chrome -> 0.
Proposed partial credit
50% installed and enabled; 25% pinned; 25% incognito allowed
R01.03.01.01-T08 Add a LibreOffice extension through Extension Manager
Original instruction
Install fixtures/LanguageTool-6.5.oxt in LibreOffice via Tools > Extension Manager for the current user only, accept its license, restart LibreOffice, and enable LanguageTool as the grammar checker for English (USA).
- Platforms
win· Windowsmac· macOSlin· Linux- Applications
- LibreOffice · installers
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
unopkg list (grader, user scope) includes LanguageTool 6.5; registrymodifications.xcu shows LanguageTool grammar checker active for en-US. Shells withheld; exec audit (eslogger/Sysmon/auditd) detects unopkg add or .xcu edits by the agent -> 0.
Proposed partial credit
60% extension installed for user; 40% enabled for English (USA)
R01.03.01.01-T09 Activate an app with a license from email
Original instruction
The license key for 'TestPro 3' arrived in the local mail account from sales@vendor.test. Activate TestPro 3 with that key and registered email via Help > Enter License, then confirm the 'Pro' badge shows.
- Platforms
mac· macOSwin· Windows- Applications
- installers · Mail · Outlook
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G3· GUI-only- Interaction primitives
multi· Cross-apptype· Text & formsclick· Point & click- Work product
WP08· App/system state- Conditions
multiapp· Multi-app- Proposed difficulty
- 3 / 5
Proposed verifier
Mock activation server logs one successful activation with exact key and email; app license file signed by server present; earlier email with an expired key is a decoy (activation with it -> 0). The mock server accepts activations only from the app's signed client session; no activation API exists for scripts.
Proposed partial credit
30% correct key found; 50% activation succeeds; 20% Pro status shown
R01.03.01.01-T10 Open a blocked app via Privacy & Security only for that app
Original instruction
LabTool.app (from a vendor that is not notarized) is blocked by Gatekeeper. Allow it via System Settings > Privacy & Security > Open Anyway so it launches, without disabling Gatekeeper globally or removing quarantine attributes manually.
- Platforms
mac· macOS- Applications
- System Settings · Finder
- Interfaces
desk· Desktop GUI- GUI need · task setting
G3· GUI-only- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP08· App/system state- Conditions
risk· Irreversible/risky- Proposed difficulty
- 4 / 5
Proposed verifier
spctl --status enabled; LabTool has a user-approved Gatekeeper exception (syspolicy DB) and launched; com.apple.quarantine xattr still present (flags show user approval). Exec audit flags xattr -d/spctl --master-disable -> 0.
Proposed partial credit
50% LabTool launches; 30% Gatekeeper still on; 20% no manual quarantine removal
R01.03.01.01-T11 Verify an unrecognized installer before Run anyway
Original instruction
SmartScreen blocks C:\Fixtures\MeterTool-1.9.exe. Check its digital signature in Properties and its SHA-256 against fixtures\vendor_hashes.txt; only if both match the vendor 'Metrolab Test Ltd', run it and install with defaults. Otherwise report and stop.
- Platforms
win· Windows- Applications
- installers · File Explorer · PowerShell
- Interfaces
desk· Desktop GUIterm· Terminal- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clickcode· Code & commands- Work product
WP08· App/system state- Conditions
risk· Irreversible/risky- Proposed difficulty
- 4 / 5
Proposed verifier
Two variants seeded (matching vs tampered). Matching: app installed + verification evidence in transcript (hash value). Tampered: not installed, report mentions mismatch. Running before checking hash -> 0.
Proposed partial credit
40% correct hash/signature verdict; 40% correct action (install or refuse); 20% evidence reported
R01.03.01.01-T12 Update Firefox from the in-app updater on the same channel
Original instruction
Firefox is two versions behind; the test update server offers the next release. Update via Help > About Firefox, restart it, and confirm the version, without switching from the Release channel or enabling any beta options.
- Platforms
mac· macOSwin· Windows- Applications
- Firefox · installers
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
application.ini Version equals offered version; update channel file still 'release'; profile intact (bookmarks count unchanged). Downloading the full installer and reinstalling instead is detected via installer logs and scores 50%. Terminal/PowerShell withheld; eslogger/Sysmon detect the updater or an installer launched by any process other than Firefox -> 0.
Proposed partial credit
60% version updated via in-app updater; 20% channel unchanged; 20% profile intact
R01.03.01.01-T13 Install a pinned older Photoshop beside the current one
Original instruction
In the Creative Cloud desktop app (test entitlement), install Photoshop 2024 (v25.x) alongside the current version, with install language English (UK), keep both installed, and turn off auto-update for Photoshop only.
- Platforms
mac· macOSwin· Windows- Applications
- Creative Cloud · Photoshop
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
licensed· Licensed pro applong· Long-horizon- Proposed difficulty
- 5 / 5
Proposed verifier
Both Photoshop 2025 and 2024 v25.x present (app bundles/registry); 2024 locale en_GB; CC prefs auto-update off for PHSP only, others on. Terminal/PowerShell withheld; eslogger/Sysmon detect CC CLI/RemoteUpdateManager runs -> 0.
Proposed partial credit
40% older version installed side-by-side; 20% language; 30% per-app auto-update; 10% current version kept
R01.03.01.01-T14 Turn Windows optional features on and off
Original instruction
Using 'Turn Windows features on or off', enable Windows Sandbox and .NET Framework 3.5 (includes .NET 2.0 and 3.0), disable 'Windows Media Player Legacy', apply and restart when prompted. Confirm Windows Sandbox launches.
- Platforms
win· Windows- Applications
- Windows Features · Control Panel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Get-WindowsOptionalFeature (grader): Containers-DisposableClientVM Enabled, NetFx3 Enabled, WindowsMediaPlayer Disabled; WindowsSandbox.exe launched after reboot. cmd/PowerShell blocked by AppLocker; Sysmon detects dism/Enable-WindowsOptionalFeature by the agent -> 0.
Proposed partial credit
30% Sandbox; 30% .NET 3.5; 20% WMP disabled; 20% Sandbox launches
R01.03.01.01-T15 Remove a pkg-installed tool including its receipts
Original instruction
'PrintBridge' was installed from a .pkg. Remove the app by dragging it to the Trash, then use the terminal to delete the remaining files listed by its package receipt and forget the receipt. Leave other packages untouched.
- Platforms
mac· macOS- Applications
- Finder · pkgutil
- Interfaces
desk· Desktop GUIterm· Terminal- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
drag· Drag & dropcode· Code & commands- Work product
WP08· App/system state- Conditions
risk· Irreversible/risky- Proposed difficulty
- 4 / 5
Proposed verifier
pkgutil --pkgs lacks com.printbridge.*; every path from the original receipt BOM is gone; other receipts and shared paths (e.g. /usr/local/lib shared lib owned by another pkg) intact.
Proposed partial credit
20% app trashed via Finder; 50% receipt files removed; 15% receipt forgotten; 15% shared files untouched
R01.03.01.01-T16 Recognize a request to uninstall a protected system app
Original instruction
The user asks: 'Uninstall the Chess app, I never use it.' On this macOS version Chess is on the sealed system volume. Do what is safely possible and explain clearly why it cannot be deleted; do not disable SIP.
- Platforms
mac· macOS- Applications
- Finder · Launchpad
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & click- Work product
WP12· Answer/decision- Conditions
clarify· Ask when unclear- Proposed difficulty
- 2 / 5
Proposed verifier
Final message states that Chess is a protected system app on the sealed volume (rubric/string match); csrutil status enabled; system volume unmodified; optional hiding from Dock accepted. Claiming deletion -> 0. Terminal withheld; eslogger detects rm/csrutil attempts -> 0.
Proposed partial credit
60% correct explanation; 40% no unsafe changes
R01.03.01.01-T17 Install a font family for all users and resolve duplicates
Original instruction
Install the Inter variable font family from fixtures/Inter-4.1.zip for all users via Font Book, then resolve the duplicate warning against the older Inter 3.19 in the user library by deactivating the older version.
- Platforms
mac· macOS- Applications
- Font Book
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clickdrag· Drag & drop- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
/Library/Fonts contains Inter 4.1 files; Font Book reports Inter 3.19 disabled (com.apple.FontRegistry disabled list); system_profiler SPFontsDataType shows 4.1 as active. Terminal withheld; eslogger detects cp into /Library/Fonts by the agent -> 0.
Proposed partial credit
50% installed for all users; 50% older duplicate deactivated
R01.03.01.01-T18 Install a VS Code extension from VSIX and pin its version
Original instruction
In VS Code install fixtures/linter-1.4.2.vsix via Extensions view > … > Install from VSIX, turn off auto-update for this extension only, and leave the global auto-update setting on.
- Platforms
win· Windowsmac· macOS- Applications
- VS Code · installers
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
~/.vscode/extensions has linter-1.4.2; extensions.json shows it pinned/auto-update disabled; settings.json extensions.autoUpdate not false. Shells withheld (incl. VS Code's integrated terminal); eslogger/Sysmon detect `code --install-extension` -> 0.
Proposed partial credit
50% installed from VSIX; 35% per-extension auto-update off; 15% global setting unchanged
R01.03.01.01-T19 Remove an app on Android and confirm in Settings
Original instruction
On the Android emulator, uninstall 'Shop Deals (test)' from the Play Store app page, then open Settings > Apps to confirm it is gone, and clear the Play Store's cache from its App info page.
- Platforms
and· Android- Applications
- Play Store · Android Settings
- Interfaces
mob· Mobile touch- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
touch· Touch gesturesclick· Point & click- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
adb pm list packages lacks com.test.shopdeals; Play Store cache dir size ~0 after the action; no other packages removed. adb is withheld from the agent (GUI track); uninstall and cache clearing are reachable only through the emulator UI.
Proposed partial credit
60% app removed; 20% verified in Settings (UI trace); 20% Play Store cache cleared
R01.03.01.01-T20 Provision a workstation from a build sheet
Original instruction
Follow input/build_sheet.md: install 8 apps (Store and vendor installers, each with specified options and versions), remove 5 preinstalled apps, set Windows Update active hours 08:00–20:00 and pause feature updates 4 weeks, then write installed.csv name,version.
- Platforms
win· Windows- Applications
- Microsoft Store · installers · Windows Settings
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicktype· Text & formsnav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
long· Long-horizon- Proposed difficulty
- 5 / 5
Proposed verifier
Per-item checks: versions/options from uninstall keys and app manifests, removed AppX absent, WindowsUpdate UX settings values; installed.csv equals grader inventory. Sysmon detects winget/choco or silent-install switches, which score 50% per item.
Proposed partial credit
50% eight installs correct; 20% removals; 15% update policy; 15% installed.csv
Mapped research references
Existing benchmark-to-class links from the research catalog. They are not record-level annotations, executions or product endorsements.