Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Analyze Windows Shellbag registry artifacts to reconstruct folder browsing activity, detect access to removable media and network shares, and establish user interaction with directories even after deletion using SBECmd and ShellBags Explorer.
.claude/skills/analyzing-windows-shellbag-artifacts/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | — | — |
| case-13 | ✗→✓ | ▲ Improved | — | — |
| case-09 | ✗→✓ | ▲ Improved | — | — |
| case-14 | ✗→✓ | ▲ Improved | — | — |
| case-17 | ✗→✗ | = Same ✗ | — | — |
Shellbags are Windows registry artifacts that track how users interact with folders through Windows Explorer, storing view settings such as icon size, window position, sort order, and view mode. From a forensic perspective, Shellbags provide definitive evidence of folder access -- even folders that no longer exist on the system. When a user browses to a folder via Windows Explorer, the Open/Save dialog, or the Control Panel, a Shellbag entry is created or updated in the user's registry hive. These entries persist after folder deletion, drive disconnection, and even across user profile resets, making them invaluable for proving that a user navigated to specific directories on local drives, USB devices, network shares, or zip archives.
| Hive | Key Path | Stores | |------|---------|--------| | NTUSER.DAT | Software\Microsoft\Windows\Shell\BagMRU | Folder hierarchy tree | | NTUSER.DAT | Software\Microsoft\Windows\Shell\Bags | View settings per folder | | UsrClass.dat | Local Settings\Software\Microsoft\Windows\Shell\BagMRU | Desktop/Explorer shell | | UsrClass.dat | Local Settings\Software\Microsoft\Windows\Shell\Bags | Additional view settings |
The BagMRU key contains a hierarchical tree of numbered subkeys representing the directory structure. Each subkey value contains a Shell Item (SHITEMID) binary blob encoding the folder identity:
Each Shell Item contains:
powershell# Parse shellbags from a directory of registry hives SBECmd.exe -d "C:\Evidence\Registry" --csv C:\Output --csvf shellbags.csv # Parse from a live system (requires admin) SBECmd.exe --live --csv C:\Output --csvf live_shellbags.csv # Key output columns: # AbsolutePath - Full reconstructed path # CreatedOn - When the folder was first browsed # ModifiedOn - When view settings were last changed # AccessedOn - Last access timestamp # ShellType - Type of shell item (Directory, Drive, Network, etc.) # Value - Raw shell item data
powershell# Launch GUI tool for interactive analysis ShellBagsExplorer.exe # Load registry hives: File > Load Hive # Navigate the tree structure to see folder hierarchy # Right-click entries for detailed shell item properties
textShellbag Path: My Computer\E:\Confidential\Project_Files ShellType: Directory (on removable volume) CreatedOn: 2025-03-15 09:30:00 UTC This proves the user navigated to E:\Confidential\Project_Files via Windows Explorer, even if the USB drive is no longer connected. The volume letter E: and directory timestamps can be correlated with USBSTOR and MountPoints2 registry entries.
textShellbag Path: \\FileServer01\Finance\Q4_Reports ShellType: Network Location AccessedOn: 2025-02-20 14:15:00 UTC This proves the user browsed to a network share, even if the share has been decommissioned or access revoked.
textShellbag Path: C:\Users\suspect\Documents\Exfiltration_Staging ShellType: Directory CreatedOn: 2025-01-10 08:00:00 UTC Even though C:\Users\suspect\Documents\Exfiltration_Staging no longer exists, the Shellbag entry proves the user created and navigated to this folder.
text$ SBECmd.exe -d "C:\Evidence\Users\jsmith" --csv /analysis/shellbag_output SBECmd v2.1.0 - ShellBags Explorer (Command Line) ==================================================== Processing hives for user: jsmith NTUSER.DAT: C:\Evidence\Users\jsmith\NTUSER.DAT UsrClass.dat: C:\Evidence\Users\jsmith\AppData\Local\Microsoft\Windows\UsrClass.dat [+] NTUSER.DAT shellbag entries: 456 [+] UsrClass.dat shellbag entries: 1,234 [+] Total shellbag entries: 1,690 --- Folder Access Timeline (Incident Window) --- Last Accessed (UTC) | Folder Path | Type | Access Count ------------------------|---------------------------------------------------------|-------------|------------- 2024-01-15 14:34:05 | C:\Users\jsmith\Downloads | File System | 45 2024-01-15 14:36:25 | C:\ProgramData\Updates | File System | 3 2024-01-15 15:05:00 | \\FILESERV01\Finance | Network | 2 2024-01-15 15:12:30 | \\FILESERV01\Finance\Q4_Reports | Network | 1 2024-01-15 15:30:00 | E:\ | Removable | 4 2024-01-15 15:30:45 | E:\Backup | Removable | 3 2024-01-15 15:31:20 | E:\Backup\Corporate_Data | Removable | 2 2024-01-15 16:12:45 | \\FILESERV01\HR\Employees | Network | 1 2024-01-15 16:15:00 | \\FILESERV01\HR\Employees\Records_2024 | Network | 1 2024-01-16 02:35:00 | C:\Windows\Temp | File System | 5 2024-01-17 02:44:00 | C:\ProgramData\svc | File System | 2 2024-01-18 01:10:00 | C:\Users\jsmith\AppData\Local\Temp | File System | 8 --- Network Share Access --- \\FILESERV01\Finance First: 2023-09-10 Last: 2024-01-15 \\FILESERV01\Finance\Q4_Reports First: 2024-01-15 Last: 2024-01-15 (NEW) \\FILESERV01\HR\Employees First: 2024-01-15 Last: 2024-01-15 (NEW) \\DC01\SYSVOL First: 2023-03-15 Last: 2024-01-16 (anomalous access time) --- Removable Device Access --- E:\ (USB Drive) Volume Name: BACKUP_DRIVE First Accessed: 2024-01-15 15:30:00 UTC Last Accessed: 2024-01-15 15:45:22 UTC Folders Browsed: 3 (E:\, E:\Backup, E:\Backup\Corporate_Data) --- Deleted/No Longer Existing Paths --- C:\ProgramData\Updates\ (folder deleted, shellbag persists) C:\ProgramData\svc\ (folder deleted, shellbag persists) C:\Windows\Temp\tools\ (folder deleted, shellbag persists) Summary: Total unique folders accessed: 1,690 Network shares accessed: 4 (2 newly accessed during incident) Removable media: 1 USB device (data staging suspected) Deleted folder evidence: 3 paths (anti-forensics indicator) CSV exported to: /analysis/shellbag_output/
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +17 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.