Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
3,324 detections
Filters
Last updated
All Time
Detection languages
1,225
935
481
239
184
Contributors
626
512
469
337
261
Categories
20,445
11,833
5,875
5,151
4,864
3,324
Platforms
1,411
1,220
873
142
138
Products / Services
973
444
393
392
360
MITRE Techniques
1,525
1,085
885
593
540
CVEs
42
28
26
26
24
IDS Classtypes
390
220
127
40
26
IDS Protocols
671
150
30
18
12
Detects ctfmon.exe calling NtCreateSymbolicLinkObject or loading cldapi.dll, which indicates the initial stages of the GreenPlasma exploit. This behavior is indicative of process injection or privilege escalation attempts.
This rule detects a potential privilege escalation attempt on Linux systems. It looks for two main indicators: first, the execution of the 'su' command from a user's home directory by a process whose executable was recently created in a user's home directory and identified as an ELF binary. Second, it identifies systems where public-facing network connections were made by processes using common remote file transfer utilities (curl, wget, lftp, axel, aria2, scp, rsync, ftp) to fetch code, followed by the execution of a process with 'algif-aead' in its command line, which could indicate fileless execution after remote code fetching. The combination suggests an attacker might be fetching malicious code, potentially compiling it, and then using it to escalate privileges via 'su' or executing it filelessly.
https://www.microsoft.com/en-us/security/blog/2026/05/08/active-attack-dirty-frag-linux-vulnerability-expands-post-compromise-risk/
Detects the execution of the public Dirty Frag exploit tool by monitoring for its hardcoded command-line arguments (--force-esp, --force-rxrpc). This vulnerability allows local attackers to gain root privileges via page-cache tampering in the Linux kernel without race conditions.
This rule detects exploitation attempts targeting CVE-2026-0300 in Palo Alto Networks PAN-OS. It identifies suspicious network connections to User-ID Authentication Portal ports (6081, 6082) from external IP addresses, excluding RFC1918 and loopback ranges. Additionally, it looks for a specific anomalous User-Agent string associated with the CL-STA-1132 threat actor, which has been observed during exploitation of this vulnerability.
Detects the addition of malicious RxRPC v1 tokens to the process keyring. This behavior is indicative of the RxRPC variant of the Dirty Frag Linux local privilege escalation exploit. The exploit uses the add_key syscall to plant a session key used for in-place pcbc(fcrypt) decryption, targeting privileged files like /etc/passwd.
Detects write operations or modifications to 'uid_map', 'gid_map', or 'setgroups' files within the /proc directory.
The xfrm-ESP variant of the Dirty Frag (CVE-2026-43284) vulnerability relies on creating new user and network namespaces, and mapping user/group IDs to root (0) to gain CAP_NET_ADMIN privileges. This setup phase allows the exploit to bypass namespace restrictions and register malicious XFRM Security Associations.
Note: Traditional auditd watches cannot monitor the /proc pseudo-filesystem directly. This detection relies on EDRs or eBPF-based sensors that can hook into file operations.
The xfrm-ESP variant of the Dirty Frag (CVE-2026-43284) vulnerability relies on creating new user and network namespaces, and mapping user/group IDs to root (0) to gain CAP_NET_ADMIN privileges. This setup phase allows the exploit to bypass namespace restrictions and register malicious XFRM Security Associations.
Note: Traditional auditd watches cannot monitor the /proc pseudo-filesystem directly. This detection relies on EDRs or eBPF-based sensors that can hook into file operations.
Detects the addition of RxRPC tokens to the process keyring using the `add_key` system call or `keyctl` command.
The RxRPC variant of the Dirty Frag (CVE-2026-43500) vulnerability leverages malicious RxRPC v1 tokens to perform an offline brute-force attack and achieve local privilege escalation via a page-cache write.
Monitoring for unexpected RxRPC key additions helps identify potential exploitation of this vulnerability.
The RxRPC variant of the Dirty Frag (CVE-2026-43500) vulnerability leverages malicious RxRPC v1 tokens to perform an offline brute-force attack and achieve local privilege escalation via a page-cache write.
Monitoring for unexpected RxRPC key additions helps identify potential exploitation of this vulnerability.
Detects a process modifying multiple privileged namespace mapping files (uid_map, gid_map, setgroups) under /proc. This behavior is used by the Dirty Frag xfrm-ESP exploit variant to gain CAP_NET_ADMIN privileges prior to exploiting the Linux kernel page-cache write vulnerability.
Detects attempts to bind an AF_ALG socket to the pcbc(fcrypt) cryptographic cipher. The Dirty Frag (CVE-2026-43500) RxRPC exploit variant leverages this highly unusual cryptographic combination to perform offline key brute-forcing in user-space to predict decrypted file offsets. This rule covers eBPF telemetry arguments, standard command-line logging (to catch potential kernel `modprobe` triggers), and auditd SOCKADDR hex-encoded structures.
Detects the malformed User-Agent string used by CL-STA-1132 during exploitation
of CVE-2026-0300 (PAN-OS Captive Portal buffer overflow). The UA is structurally
invalid (Safari token precedes Mozilla token, Mozilla/5.5 is non-standard) and
has no known legitimate origin.
of CVE-2026-0300 (PAN-OS Captive Portal buffer overflow). The UA is structurally
invalid (Safari token precedes Mozilla token, Mozilla/5.5 is non-standard) and
has no known legitimate origin.
Detects suspicious command execution originating from cPanel daemons. This rule identifies instances where cPanel processes (cpsrvd, cpsrvd-ssl, cpaneld, whostmgrd) or their immediate children execute common shell interpreters or specific commands indicative of potential compromise, such as writing to /etc/passwd, modifying authorized_keys, or using base64/perl/python/bash for interactive shells.
Detects truncated 'whostmgrsession' cookies missing the comma delimiter, which is a necessary step for attackers to inject cleartext CRLF payloads into cPanel session files and bypass authentication. This rule specifically looks for the 'whostmgrsession' cookie in web or proxy logs where the cookie value matches a regex pattern indicating the absence of the expected comma delimiter, a characteristic of the CVE-2026-41940 exploitation.
This rule detects the execution of the `unshare` utility with parameters to create a new user namespace (`-U` or `--user`). This activity is a known precursor for the Dirty Frag Linux kernel Local Privilege Escalation (LPE) exploit, where an unprivileged user can gain elevated privileges. The rule specifically targets Linux process execution events.
This rule detects the execution of the `unshare` utility with parameters to create a new user namespace (`-U` or `--user`). This activity is a known precursor for the Dirty Frag Linux kernel Local Privilege Escalation (LPE) exploit, where an unprivileged user can gain elevated privileges. The rule specifically targets Linux process execution events.
Detects inbound network connections from external, non-private IP addresses to local ports 6081 and 6082. These ports are used by the PAN-OS User-ID Authentication Portal, which is actively targeted by CVE-2026-0300.
Detects Python command lines referencing a specific cryptographic string ('authencesn(hmac(sha256),cbc(aes))') known to be used in the exploitation of the 'Copy Fail' vulnerability (CVE-2026-31431). This string is highly specific and its presence in a Python command line is indicative of an attempt to exploit this particular CVE.
Detects inbound network connections from external, non-private IP addresses to local ports 6081 and 6082. These ports are used by the PAN-OS User-ID Authentication Portal, which is actively targeted by CVE-2026-0300.
Suspicious PAN-OS Captive Portal Access
Cortex XDR
Detects anomalous or repetitive access to PAN-OS Captive Portal HTTP endpoints which may indicate active scanning for CVE-2026-0300 vulnerability exposure.
Detects the creation of a new MariaDB User-Defined Function (UDF) that loads a shared object (.so or .dll) file. This technique is used in the final stage of the CVE-2026-32710 attack chain to execute arbitrary OS commands, often from suspicious temporary or public directories.
Identifies network connections targeting port 6082, commonly used by the PAN-OS User-ID portal and specifically targeted in the CVE-2026-0300 PoC exploit to deliver a buffer overflow payload.
Page 112 of 159