Georg Lauenstein
@gsecTrusted contributorCompletionist
11 followers468 downloads740 copies45 likes15,786 views
211 detections
Filters
Last updated
All Time
Detection languages
197
14
Categories
56
46
14
12
12
Platforms
76
31
6
5
3
Products / Services
13
9
8
4
4
MITRE Techniques
47
43
38
37
21
CVEs
3
3
2
2
1
Detects execution of the XMRig Monero cryptominer or a miner
launched with XMRig-specific pool and tuning arguments. In the AhsayCBS
exploitation campaign attackers dropped XMRig onto compromised backup
servers to hijack CPU resources for Monero mining, severely degrading
server performance. Detection keys on XMRig binary names and the mining
command-line grammar (stratum pool URLs, donate-level and thread-hint flags).
launched with XMRig-specific pool and tuning arguments. In the AhsayCBS
exploitation campaign attackers dropped XMRig onto compromised backup
servers to hijack CPU resources for Monero mining, severely degrading
server performance. Detection keys on XMRig binary names and the mining
command-line grammar (stratum pool URLs, donate-level and thread-hint flags).
Detects creation of a JSP/JSPX/WAR server-side script inside
an AhsayCBS installation or web application directory. In this campaign the
actors exploited critical AhsayCBS vulnerabilities and dropped webshells into
the application web directories to maintain persistent remote access. Legitimate
deployment writes scripts from the AhsayCBS updater, not ad-hoc into web roots.
an AhsayCBS installation or web application directory. In this campaign the
actors exploited critical AhsayCBS vulnerabilities and dropped webshells into
the application web directories to maintain persistent remote access. Legitimate
deployment writes scripts from the AhsayCBS updater, not ad-hoc into web roots.
Detects an agentic AI application (e.g., Claude Code, ChatGPT Codex) directly
spawning a package manager install command (npm install / pip install). As enterprise
adoption of agentic coding and automation tools spreads beyond developer workstations into
marketing, HR, finance, and operations, these tools autonomously pull open-source packages
from npm and PyPI on behalf of non-developer users. CrowdStrike reports poisoned packages
(e.g., 131 AI framework packages by STARDUST CHOLLIMA, 300+ dependencies in a single day by
ALTERED SPIDER) are increasingly pulled into environments this way, giving adversaries a path
to endpoints that never had traditional developer tooling scrutiny.
spawning a package manager install command (npm install / pip install). As enterprise
adoption of agentic coding and automation tools spreads beyond developer workstations into
marketing, HR, finance, and operations, these tools autonomously pull open-source packages
from npm and PyPI on behalf of non-developer users. CrowdStrike reports poisoned packages
(e.g., 131 AI framework packages by STARDUST CHOLLIMA, 300+ dependencies in a single day by
ALTERED SPIDER) are increasingly pulled into environments this way, giving adversaries a path
to endpoints that never had traditional developer tooling scrutiny.
Detects the PatchAgent Stage 3 HTTP C2 protocol recovered from the decrypted shellcode: plain HTTP requests to /api/checkin, /api/cmd/, or /api/result/ combined with the hardcoded, truncated User-Agent "Mozilla/5.0 (Windows NT 10.0; Win64; x64)" (a real browser UA continues past this). Matching the API path plus the exact short UA keeps this durable across C2 IP rotation (observed on 46.30.188[.]99 and 185.228.83[.]217, port 8443).
Detects creation of sitecustomize.py, usercustomize.py, or .pth path configuration files inside a site-packages or dist-packages directory by a python or pip process. These files are automatically loaded by Python's site module on every subsequent interpreter start, allowing a malicious package installer to establish persistence that re-executes on every future Python invocation without further victim interaction.
Detects an AhsayCBS / Tomcat / Java web service process spawning a
command or script interpreter. After exploiting AhsayCBS, attackers executed
shell commands through the dropped webshell to run reconnaissance and download
the XMRig payload. A backup web service should never directly spawn cmd.exe,
powershell.exe, or a Unix shell - this indicates webshell command execution.
command or script interpreter. After exploiting AhsayCBS, attackers executed
shell commands through the dropped webshell to run reconnaissance and download
the XMRig payload. A backup web service should never directly spawn cmd.exe,
powershell.exe, or a Unix shell - this indicates webshell command execution.
Detects repeated OAuth device authorization ('device code') grant flow authentication events targeting the same user within a short window, which is indicative of an active device code phishing attempt.
Detects creation of files or folders under the hard-coded staging
directory C:\Users\Public\Temp\ChromeAuto_. Astaroth's spambot
copies the victim's browser "User Data" profile directory to this path so it can
spawn a fresh browser instance carrying the victim's active WhatsApp Web session
cookies and stored browser credentials, bypassing authentication entirely.
directory C:\Users\Public\Temp\ChromeAuto_. Astaroth's spambot
copies the victim's browser "User Data" profile directory to this path so it can
spawn a fresh browser instance carrying the victim's active WhatsApp Web session
cookies and stored browser credentials, bypassing authentication entirely.
Detects a Chrome/Edge WebDriver process launched with headless and
automation-hiding command-line flags (--headless combined with --disable-infobars
or excludeSwitches enable-automation/test-type). Astaroth's spambot component uses
this exact flag combination to drive a hidden browser instance against an active
WhatsApp Web session, sending spam to every Brazil-based contact without ever
displaying a window to the victim.
automation-hiding command-line flags (--headless combined with --disable-infobars
or excludeSwitches enable-automation/test-type). Astaroth's spambot component uses
this exact flag combination to drive a hidden browser instance against an active
WhatsApp Web session, sending spam to every Brazil-based contact without ever
displaying a window to the victim.
Detects cmd.exe spawned by a python.exe or pip.exe process with a command line invoking netstat or calc.exe. Matches the payload behavior described for the malicious redpy_demo PyPI package, whose __main__.py used Python's subprocess module to execute these commands via cmd.exe automatically at module invocation, without any direct victim interaction with the payload.
Page 6 of 22