Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Analyze Windows LNK shortcut files and Jump List artifacts to establish evidence of file access, program execution, and user activity using LECmd, JLECmd, and manual binary parsing of the Shell Link Binary format.
.claude/skills/analyzing-lnk-file-and-jump-list-artifacts/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | — | — |
| case-10 | ✗→✓ | ▲ Improved | — | — |
| case-24 | ✗→✓ | ▲ Improved | — | — |
| case-25 | ✗→✓ | ▲ Improved | — | — |
| case-11 | ✗→✓ | ▲ Improved | — | — |
Windows LNK (shortcut) files and Jump Lists are critical forensic artifacts that provide evidence of file access, program execution, and user behavior. LNK files are created automatically when a user opens a file through Windows Explorer or the Open/Save dialog, storing metadata about the target file including its original path, timestamps, volume serial number, NetBIOS name, and MAC address of the host system. Jump Lists, introduced in Windows 7, extend this by maintaining per-application lists of recently and frequently accessed files. These artifacts persist even after the target files are deleted, making them invaluable for establishing that a user accessed specific files at specific times.
| Location | Description | |----------|-------------| | %USERPROFILE%\AppData\Roaming\Microsoft\Windows\Recent\ | Recent files accessed | | %USERPROFILE%\Desktop\ | User-created shortcuts | | %USERPROFILE%\AppData\Roaming\Microsoft\Windows\Start Menu\ | Start Menu shortcuts | | %USERPROFILE%\AppData\Roaming\Microsoft\Office\Recent\ | Office recent documents |
| Offset | Size | Field | |--------|------|-------| | 0x00 | 4 | HeaderSize (always 0x0000004C) | | 0x04 | 16 | LinkCLSID (always 00021401-0000-0000-C000-000000000046) | | 0x14 | 4 | LinkFlags | | 0x18 | 4 | FileAttributes | | 0x1C | 8 | CreationTime (FILETIME) | | 0x24 | 8 | AccessTime (FILETIME) | | 0x2C | 8 | WriteTime (FILETIME) | | 0x34 | 4 | FileSize of target | | 0x38 | 4 | IconIndex | | 0x3C | 4 | ShowCommand | | 0x40 | 2 | HotKey |
powershell# Parse all LNK files in Recent folder LECmd.exe -d "C:\Evidence\Users\suspect\AppData\Roaming\Microsoft\Windows\Recent" --csv C:\Output --csvf lnk_analysis.csv # Parse a single LNK file with full details LECmd.exe -f "C:\Evidence\Users\suspect\Desktop\Confidential.docx.lnk" --json C:\Output # Parse LNK files with additional detail levels LECmd.exe -d "C:\Evidence\Users\suspect\AppData\Roaming\Microsoft\Windows\Recent" --csv C:\Output --csvf lnk_all.csv --all
powershell# Parse Automatic Jump Lists JLECmd.exe -d "C:\Evidence\Users\suspect\AppData\Roaming\Microsoft\Windows\Recent\AutomaticDestinations" --csv C:\Output --csvf jumplists_auto.csv # Parse Custom Jump Lists JLECmd.exe -d "C:\Evidence\Users\suspect\AppData\Roaming\Microsoft\Windows\Recent\CustomDestinations" --csv C:\Output --csvf jumplists_custom.csv # Parse all jump lists with detailed output JLECmd.exe -d "C:\Evidence\Users\suspect\AppData\Roaming\Microsoft\Windows\Recent\AutomaticDestinations" --csv C:\Output --csvf jumplists_auto.csv --ld
These are OLE Compound files (Structured Storage) identified by AppID hash in the filename:
| AppID Hash | Application | |-----------|-------------| | 5f7b5f1e01b83767 | Windows Explorer Pinned/Frequent | | 1b4dd67f29cb1962 | Windows Explorer Recent | | 9b9cdc69c1c24e2b | Notepad | | a7bd71699cd38d1c | Notepad++ | | 12dc1ea8e34b5a6 | Microsoft Paint | | 7e4dca80246863e3 | Control Panel | | 1cf97c38a5881255 | Microsoft Edge | | f01b4d95cf55d32a | Windows Explorer | | 9d1f905ce5044aee | Microsoft Excel | | a4a5324453625195 | Microsoft Word | | d00655d2aa12ff6d | Microsoft PowerPoint | | bc03160ee1a59fc1 | Outlook |
Created when users pin items to application jump lists. These files contain sequential LNK entries.
pythonimport struct import os from datetime import datetime, timedelta FILETIME_EPOCH = datetime(1601, 1, 1) def filetime_to_datetime(filetime_bytes: bytes) -> datetime: """Convert Windows FILETIME (100-ns intervals since 1601) to datetime.""" ft = struct.unpack("<Q", filetime_bytes)[0] if ft == 0: return None return FILETIME_EPOCH + timedelta(microseconds=ft // 10) def parse_lnk_header(lnk_path: str) -> dict: """Parse the Shell Link header from an LNK file.""" with open(lnk_path, "rb") as f: header = f.read(76) header_size = struct.unpack("<I", header[0:4])[0] if header_size != 0x4C: return {"error": "Invalid LNK header"} link_flags = struct.unpack("<I", header[0x14:0x18])[0] file_attrs = struct.unpack("<I", header[0x18:0x1C])[0] result = { "header_size": header_size, "link_flags": hex(link_flags), "file_attributes": hex(file_attrs), "creation_time": filetime_to_datetime(header[0x1C:0x24]), "access_time": filetime_to_datetime(header[0x24:0x2C]), "write_time": filetime_to_datetime(header[0x2C:0x34]), "file_size": struct.unpack("<I", header[0x34:0x38])[0], "has_target_id_list": bool(link_flags & 0x01), "has_link_info": bool(link_flags & 0x02), "has_name": bool(link_flags & 0x04), "has_relative_path": bool(link_flags & 0x08), "has_working_dir": bool(link_flags & 0x10), "has_arguments": bool(link_flags & 0x20), "has_icon_location": bool(link_flags & 0x40), } return result
Recent research (IEEE 2025) shows that Windows 11 produces different LNK and Jump List artifacts:
text$ LECmd.exe -d "C:\Evidence\Users\jsmith\AppData\Roaming\Microsoft\Windows\Recent" --csv /analysis/lnk_output LECmd v1.11.0 - LNK File Parser ================================ Processing 47 LNK files... --- LNK File: Q4_Report.xlsx.lnk --- Source: C:\Evidence\Users\jsmith\Recent\Q4_Report.xlsx.lnk Target Path: C:\Users\jsmith\Downloads\Q4_Report.xlsm Target Created: 2024-01-15 14:33:45 UTC Target Modified: 2024-01-15 14:33:45 UTC Target Accessed: 2024-01-15 14:35:12 UTC File Size: 251,904 bytes Drive Type: Fixed (C:) Volume Serial: A4E7-3F21 Machine ID: DESKTOP-J5M1TH MAC Address: 48:2A:E3:5C:9B:01 --- LNK File: update_client.exe.lnk --- Source: C:\Evidence\Users\jsmith\Recent\update_client.exe.lnk Target Path: C:\ProgramData\Updates\update_client.exe Target Created: 2024-01-15 14:34:02 UTC Target Modified: 2024-01-15 14:34:02 UTC Target Accessed: 2024-01-15 14:36:30 UTC File Size: 1,258,496 bytes Drive Type: Fixed (C:) Volume Serial: A4E7-3F21 Machine ID: DESKTOP-J5M1TH Working Dir: C:\ProgramData\Updates Arguments: --silent --no-update-check Run Window: Hidden ====================================================================== $ JLECmd.exe -d "C:\Evidence\Users\jsmith\AppData\Roaming\Microsoft\Windows\Recent\AutomaticDestinations" --csv /analysis/jumplist_output JLECmd v1.5.0 - Jump List Parser ================================== Processing 23 AutomaticDestinations files... --- Application: Microsoft Excel (AppID: 12dc1ea8e34b5a6) --- Entries: 15 Most Recent: Entry 0: C:\Users\jsmith\Downloads\Q4_Report.xlsm (2024-01-15 14:35:12 UTC) Entry 1: \\FILESERV01\Finance\Budget_2024.xlsx (2024-01-14 09:22:30 UTC) Entry 2: C:\Users\jsmith\Documents\Expenses\Dec2023.xlsx (2024-01-10 16:45:00 UTC) --- Application: Windows Explorer (AppID: f01b4d95cf55d32a) --- Entries: 28 Most Recent: Entry 0: C:\ProgramData\Updates\ (2024-01-15 14:36:25 UTC) Entry 1: E:\Backup\ (2024-01-15 15:30:00 UTC) Entry 2: \\FILESERV01\HR\Employees\ (2024-01-15 16:12:45 UTC) --- Application: cmd.exe (AppID: 9b9cdc69c1c24e2b) --- Entries: 5 Most Recent: Entry 0: C:\Windows\System32\cmd.exe (2024-01-15 14:36:00 UTC) Summary: Total LNK files processed: 47 Total Jump List entries: 156 Suspicious artifacts: 3 (hidden window execution, USB drive access, network shares) CSV exported to: /analysis/lnk_output/ and /analysis/jumplist_output/
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-08 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-24 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-26 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-25 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
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. 26 cases were attempted, and 25 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +23 percentage points is the difference between those two pass rates over the 25 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.