
Passkey Help-Desk Vishing: Rogue MFA Registration to SharePoint Data Theft
An identity alert fired at 13:53 UTC on 29 September on Tomas Ferreira's account at Northcape Mutual Insurance: a new Microsoft Authenticator method was registered. The service desk is halfway through a company passkey rollout (CHG-2026-0412), three identical alerts were closed as benign earlier in the wave, and Tomas's own note on the ticket says IT called him that afternoon about the rollout, walked him through the enrolment page and that he approved the prompt on his phone. A second alert for a bulk download on the same account landed in the queue that evening, next to one for a Claims colleague that was closed against a change record three hours earlier. You have three weeks of Entra sign-in and directory audit events, the Microsoft 365 activity for SharePoint, OneDrive and Exchange, the estate's Windows and web telemetry and the SOC's own case history. Work the evidence to establish what was added to the account, from where, and what happened next.
Start this operation
Investigation Tasks
Complete each task by investigating alerts and submitting your findings.
Where the newcomer came from
5The case opened on an identity alert for the employee named in the briefing. Open the identity events in the panel and lay out three weeks of this employee's logons: the places the managed laptop, the enrolled phone and the usual networks come from. On the afternoon of 29 September a web logon for the same employee arrived from a place that history does not contain. Submit the IP it came from.
What the alert is really about
5The briefing says a second way of approving prompts was added to the same employee on that afternoon, and the ticket note says it was a passkey done with IT on the line. The identity events in the panel keep the record that added it, written by the platform itself rather than by the employee. Find that record and submit the tag the newly added entry carries.
The page he was read
10While the caller was still on the line, the employee typed the address the caller read out into the browser on his laptop. The web and name-resolution telemetry hold what that address was and where it led, a couple of minutes before the outside logon from the first task. Submit the full host name he typed.
The other heavy reader
5Two identities pulled far more items than usual out of the libraries on the afternoon of 29 September, and both tripped the same detection. The second case landed that evening on tomas.ferreira; the first had been raised on a colleague and settled hours earlier. Submit where the settled case's closing comment says the colleague's items were being moved to, written as the comment writes it.
What did the pulling
10Return to the identity from the first task. Between that afternoon and the following morning, dozens of items left its libraries and its own storage, retrieved from the two outside origins rather than from the laptop. Identify the software that did the retrieving and submit the agent string exactly as the records carry it.
The morning after
5Before seven the next morning the same identity logged in once more from outside, from a second origin, and the second check was cleared by typing a one-time value rather than by approving a push. Submit the name of that check as the record labels it, not its longer description, exactly as written.
How much of the mailbox was read
10The libraries were not the only thing read from outside. Over the afternoon of 29 September and again the next morning, the identity's mailbox was read through the Exchange API from the two outside origins, in batches rather than message by message. Privacy and legal need the size of the exposure before they decide on notification. Submit the total number of mailbox items those reads covered across both days, as a whole number.
7 tasks · 50 points total
Training Tools
SIEM Console
Log analysis & SPL queries
Cloud Console
Cloud infrastructure logs
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 Cloud concepts
Ready to investigate?
More Operations
View allBuilding a Backdoor
Velodraft Systems runs its release pipeline on a self-hosted TeamCity server that was left accessible from the internet on an unpatched build. An attacker exploited an authentication-bypass flaw to create a rogue administrator without credentials, harvested stored VCS secrets and cloud keys, then planted a malicious build step in the production release pipeline. When the next scheduled build fired, both CI build agents pulled and executed an implant that beaconed out to an attacker-controlled domain. Reconstruct the intrusion chain from the TeamCity audit trail and the build-agent XDR data.
The Gentlemen Ransomware: FortiVPN Login to EDR Kill and Rclone Exfiltration
Late one evening the endpoint sensor on a management server went quiet after a run of administrative-looking activity: a VPN login, a remote session, a new tunnel service, a scan exclusion and a large outbound transfer. Internal IT reaches servers from the VPN pool, runs the same tunnel product elsewhere and keeps a documented exclusion, so at first glance much of this looks routine. Work the FortiGate, Windows, Sysmon, endpoint and firewall evidence and answer each task from what the panels show.
MSP360 and ScreenConnect: Phishing Installer to Dual-RMM Persistence
Overnight hunting flagged two signed remote management agents calling home from one advisory associate's workstation, both installed within the same hour yesterday morning, with PowerShell running under one of them. The user says they opened a document review link; IT runs a sanctioned remote-support tool elsewhere, so at first read this looks like a rollout. Work the email, endpoint, SIEM and firewall evidence to establish how the agents arrived and what they were used for.