
Search Engine Poisoning: From a Sponsored Result to the Domain Controller
An analyst at a mid-sized logistics firm searched for a routine software update, clicked the sponsored result at the top of the page, and went back to work. Five days later the SOC is holding a workstation, a file server and a domain controller whose telemetry nobody has read yet. You have the SIEM and the endpoint console for those five days, and most of what is in them is an ordinary week. Reconstruct what the click actually started, follow it off the workstation, and put names to the files and the account behind the activity that does not belong.
Start this operation
Investigation Tasks
Complete each task by investigating alerts and submitting your findings.
Trace the SEO Ad to Its Callback Domain
5amanda.garcia's shift began with a click on a sponsored "Software Update" search result, and within seconds her browser walked through more than one external host. One of them appears once and hands the browser straight on; the other is the host the workstation keeps coming back to for the rest of the case. Working in the SIEM and XDR timeline around that click, name the second one. Other users' browsers reach it later in the week too, so anchor on amanda.garcia's session rather than on the domain alone.
Find What the Fake Update Page Actually Delivered
5Having established the external host that opens this case, follow what arrived next on corp-wks-442. Shortly after the browser reached that host, a file landed in amanda.garcia's profile and was opened from the shell. Using the XDR timeline's file-creation events and the SIEM process activity in the opening minutes, identify the delivered artifact that execution starts from.
Identify the Module Run from the Mounted Image
10Opening the mounted image did not run anything on its own: a shortcut inside it handed control to a signed Windows utility, which then loaded a module from a temporary directory instead of from a signed system path. Walk the execution tree spawned by the shell on corp-wks-442 and name the module the utility loaded.
Fingerprint the First Dropped Executable
10One generically named executable is written into amanda.garcia's Temp directory on corp-wks-442 and run from there four separate times across the case, each run under a different hash. To confirm scope and check the samples against threat intelligence, the team needs durable identifiers rather than the shared filename. Pivot to the process-creation events in the SIEM for that workstation, put the four runs in order, and recover the file hash of the EARLIEST one.
Spot the File Whose Extension Does Not Match Its Use
5The next day, a file with a text extension was written into amanda.garcia's Temp folder on corp-wks-442 and then appeared as the image of a newly created process. Review the XDR file-system artifacts and the SIEM process-creation events for this host and name that file.
Confirm a Second, Distinct Payload Variant
5The same Temp-folder execution pattern repeats on corp-wks-442 under amanda.garcia's account until it stops. To show the operator kept re-packing the payload rather than re-running one file, take the LAST run in that series and recover its hash from the SIEM, then confirm it differs from the earliest one you already submitted.
Attribute the Hands-on-Keyboard Phase to an Account
5As the case matures, the telemetry shifts away from amanda.garcia's browser-driven activity on the workstation and onto corp-dc-01, where a cluster of administrative tooling runs inside a few minutes. Several accounts touch the domain controller across the week, including a backup service account and two ordinary users. Work the SIEM around that cluster and determine which account was driving it.
Isolate the Odd Binary in the Final Execution Burst
5In the closing stage of the case, corp-wks-442 produced a dense burst of process executions inside twenty minutes. Most of them are ordinary Windows and office binaries. Review the XDR timeline and the matching SIEM process-creation events for this burst and identify the one executable that corresponds to no standard Windows component and no known business application.
8 tasks · 50 points total
Training Tools
Skills You'll Build
Requires foundational alert triage skills. Multiple data sources to correlate.
Prerequisites
- Basic understanding of security alerts
- Familiarity with SIEM concepts
- Familiarity with XDR concepts
Ready to investigate?
More Operations
View allCI/CD Pipeline Hijack: GitHub Actions Compromise
Investigate the March 2025 GitHub Actions supply-chain compromise involving tj-actions/changed-files and reviewdog/action-setup. A compromised action version tag caused Linux CI runners to execute malicious payload logic and expose CI/CD secrets in workflow logs using double-base64 encoding. Analyze SIEM and XDR telemetry to identify the affected action, runner identity, payload execution, detection source, and secret-exposure pattern, then decide which artifacts are malicious versus benign threat-intelligence lookups.
Edge Device Exploitation: VPN Zero-Day
Investigate a breach targeting exposed edge security appliances. In this scenario, an attacker exploits Ivanti Connect Secure CVE-2025-22457 to obtain unauthenticated code execution, spawns a shell from the gateway web process, retrieves a Linux payload, and attempts to preserve access using edge-device service abuse patterns similar to FortiGate SSL-VPN post-exploitation tradecraft. Analyze SIEM, XDR, and firewall evidence to identify the spawned shell, suspicious domain, local sync script, and defensive block policy.
Frozen Assets: The Snowflake Tenant Heist
A retailer's Snowflake tenant is accessed with a contractor credential lifted from an infostealer log. With no MFA and no network policy to stop it, the actor logs in from VPN-exit IPs, recons with anomalous tooling, and bulk-exfiltrates an entire customer table. Work the Snowflake audit trail to reconstruct the theft.