Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
257 detections
Filters
Last updated
All Time
Detection languages
117
31
27
22
16
Contributors
99
60
33
25
8
Categories
187
123
62
58
43
Platforms
39,894
7,228
6,523
3,864
3,624
257
Products / Services
146
48
47
28
22
MITRE Techniques
95
94
69
56
41
CVEs
10
5
5
5
5
IDS Classtypes
9
7
4
3
2
IDS Protocols
23
3
1
Detects the presence of the 'SOUL.md' configuration file associated with the Hermes Agent AI framework. This file contains a specific 'GH0ST' persona prompt designed to bypass ethical restrictions (jailbreak) and configure the agent for automated post-exploitation tasks, such as loot collection and Telegram-based command and control.
Detects the deployment of the CARBONATO botnet's Hermes Agent within containerized environments. The rule identifies specific artifacts such as the 'SOUL.md' persona configuration, 'netd-svc' container references, and indicators of AI-focused credential theft (e.g., searching for OpenAI or Anthropic API keys) orchestrated by the GH0ST AI agent.
Detects the use of the 'nsenter' utility to target the PID 1 namespace with host-level namespace flags. This technique is used by the CARBONATO botnet to escape from a privileged Docker container and execute commands on the underlying host system, leveraging the host filesystem mount.
This rule detects unauthorized attempts to create privileged containers via the Docker API, specifically identifying the use of the 'net-setup' string often associated with the CARBONATO worm. Such activity indicates potential propagation or lateral movement within a containerized environment by attempting to deploy containers with elevated permissions.
Detects unauthorized access to unauthenticated Docker registry API endpoints on port 5000. This rule identifies catalog enumeration and the pulling of malicious container images associated with the CARBONATO botnet, including the retrieval of image manifests and configuration files that expose sensitive environment variables and credentials.
Detects persistence mechanisms used by the CARBONATO botnet on Linux-based Docker hosts. The rule monitors for the application of immutable file attributes using 'chattr +i' on identified watchdog binaries (specifically '.docker-network-monitor'), the creation of malicious cron jobs, the enabling of suspicious systemd timers ('docker-network-resolver'), and the execution of specific persistence-related shell scripts.
Detects exploitation attempts by the CARBONATO botnet targeting exposed Docker daemons. The rule monitors for container escape techniques using 'nsenter' to access host namespaces, the execution of specific malicious entrypoint scripts, the use of known malicious container image references, and the creation of privileged containers with host bind mounts, which are indicative of host filesystem and network compromise.
Detects anomalous process execution originating from a Python-based sandbox (such as Semantic Kernel SessionsPythonPlugin in Azure Container Apps) that attempts to break out of the container isolation. The rule monitors for command-line arguments involving interaction with the Docker socket, sensitive host namespaces, or container privilege escalation techniques, indicative of exploitation attempts related to CVE-2026-25592.
Detects malicious BASH_ENV injection within GitHub Actions CI pipelines originating from the compromised sckit_poetry_build.py build backend. The injection, which targets GITHUB_ENV to introduce a malicious _pypi_bridge.sh script, is used to intercept PyPI publishing tokens during build processes, characteristic of the MemTensor/sckit supply chain attack.
Detects anomalous DNS query patterns originating from AI agent hosts that are characteristic of DNS tunneling, such as the use of TXT or NULL records and excessively long subdomain labels. These techniques are often used by attackers to establish covert Command and Control (C2) channels and exfiltrate data by bypassing traditional HTTP/HTTPS egress filtering.
Detects outbound HTTP POST requests from an AI agent runtime environment to known external paste-sites, webhook providers, or staging services. This behavior is indicative of a compromise where the AI agent is being manipulated to exfiltrate sensitive API keys, tokens, or configuration secrets to an attacker-controlled endpoint.
Detects periodic, repetitive outbound HTTPS network connections from AI agent-runtime environments to domains that are not in the predefined allow-list. This behavior is consistent with C2 beaconing disguised as application API traffic, potentially indicating an AI agent has been compromised or misused for unauthorized external command communication.
Detects host-side indicators of potential container escapes and unauthorized access, including mounting sensitive container runtime sockets, executing containers with privileged flags, host namespace sharing, and the use of nsenter to break out of a container context.
Detects attempts by AI agent code-interpreter or sandbox processes to break out of their containerized environment. The rule monitors for the execution of common container escape techniques, such as namespace manipulation (nsenter, unshare), mounting host filesystems, accessing sensitive host devices like /dev/kmsg or /dev/mem, or referencing host-level paths from within an isolated AI agent execution process.
This rule detects unauthorized tampering with AI container images within a container registry. It identifies when existing image tags are overwritten by actors outside of authorized CI/CD pipelines, or when production MLOps infrastructure pulls unsigned or invalid container images, potentially indicating a supply chain attack targeting AI model artifacts.
Detects instances where package management tools (pip, npm, docker) spawn child processes that immediately attempt to read sensitive files (SSH keys, cloud/container credentials) or perform outbound network connections. This behavior is indicative of a trojanized supply-chain package executing malicious post-install logic to exfiltrate secrets or establish initial communication with a C2 server.
Detects CARBONATO botnet activity where malicious processes are masquerading as legitimate Linux system components. This includes XMRig mining software running as /usr/sbin/systemd-logind, process arguments attempting to mimic kernel kworker threads, or processes running from directories associated with netd-svc/resolved container images.
Detects malicious interactions with an unauthenticated Docker daemon API, specifically identifying the creation of privileged containers with host bind mounts, the deployment of containers with specific naming conventions for network reconnaissance (netns-probe/net-setup), subsequent container start events, and the use of the Docker exec API to perform host namespace escapes via nsenter.
Detects anomalous activity associated with the CARBONATO worm, specifically monitoring for unauthorized creation of 'net-setup' containers or pulling 'system/resolved' images via the Docker API port 2375. Additionally, the rule identifies rapid, sequential network scanning behavior targeting the Docker API port across internal IP ranges.
Detects unauthorized access attempts to a Docker Registry API on port 5000. The rule identifies attempts to enumerate the image catalog and extract sensitive environment variable information from configuration blobs, which may contain hardcoded credentials or indicators of compromise.
Detects the execution of the XMRig cryptominer masquerading as the legitimate systemd-logind service. The rule identifies suspicious command-line arguments indicative of mining activity, or instances where systemd-logind is spawned from unauthorized parent processes such as shell interpreters, container engines, or nsenter, which is often characteristic of post-exploitation activity like privileged Docker container escapes.
Page 2 of 13