Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
1,084 detections
Filters
Last updated
All Time
Detection languages
418
193
163
156
81
Contributors
381
251
92
71
40
Categories
498
271
255
135
123
Platforms
443
441
267
121
114
Products / Services
10,665
9,758
6,692
4,418
3,749
1,084
MITRE Techniques
278
229
180
177
151
CVEs
17
15
13
13
8
IDS Classtypes
4
IDS Protocols
4
An ABAP program invoked a remote function call (RFC) using a destination resolved at runtime, which may indicate lateral movement or unauthorized internal communication within an SAP environment.
Detects multiple distinct SAP users authenticating successfully from a single source IP address within a short time window (30 minutes), potentially indicating compromised credentials or shared access anomalies.
Detects bursts of five or more distinct Cisco AI Defense event IDs, or three or more occurrences of the same event ID, matching prompt injection, code detection, PII, or PCI guardrails for the same application and connection within ten minutes.
Allowed PII/PCI prompt events are handled by Cisco AI Defense Allowed Sensitive Prompt.
These patterns can indicate iterative prompt-injection probing or repeated attempts to use an AI application to access restricted data or capabilities.
Allowed PII/PCI prompt events are handled by Cisco AI Defense Allowed Sensitive Prompt.
These patterns can indicate iterative prompt-injection probing or repeated attempts to use an AI application to access restricted data or capabilities.
The following analytic detects the behavior of decoding base64-encoded data and passing it to a Linux shell. Additionally, it mitigates the potential damage and protects the organization's systems and data.The detection is made by searching for specific commands in the Splunk query, namely "base64 -d" and "base64 --decode", within the Endpoint.Processes data model. The analytic also includes a filter for Linux shells. The detection is important because it indicates the presence of malicious activity since Base64 encoding is commonly used to obfuscate malicious commands or payloads, and decoding it can be a step in running those commands. It suggests that an attacker is attempting to run malicious commands on a Linux system to gain unauthorized access, for data exfiltration, or perform other malicious actions.
This rule monitors NetScaler ADC logs for indicators associated with the PitScaler campaign, specifically targeting exploitation of CVE-2026-88771 and CVE-2026-88772. It detects exploitation artifacts (DTLS handshake errors, NSPPE crashes), deployment of WHIPSHOT/SLAPSHOT malware, web shell installation, httpd.conf configuration tampering, persistence mechanisms involving privilege escalation (chmod), and connections to known malicious infrastructure.
Detects M365 Copilot jailbreak attempts through prompt injection techniques including rule manipulation, system bypass commands, and AI impersonation requests that attempt to circumvent built-in safety controls. The detection searches exported eDiscovery prompt logs for jailbreak keywords like "pretend you are," "act as," "rules=," "ignore," "bypass," and "override" in the Subject_Title field, assigning severity scores based on the manipulation type (score of 4 for amoral impersonation or explicit rule injection, score of 3 for entity roleplay or bypass commands). Prompts with a jailbreak score of 2 or higher are flagged, prioritizing the most severe attempts to override AI safety mechanisms through direct instruction injection or unauthorized persona adoption.
Detects high Ollama API request volumes from a single client in a short time window. A burst may indicate automated abuse or contribute to resource contention, but GIN access logs do not establish malicious intent or service impact; validate the client and workload against local usage baselines.
Identifies repeated Ollama inference requests with long response times on /api/generate, /api/chat, and /v1/chat/completions.
Long response time can result from model size, prompt length, host resources, or normal workload. GIN access logs do not include prompt content and cannot establish a prompt injection or jailbreak attempt, so use this as an exploratory performance signal.
Long response time can result from model size, prompt length, host resources, or normal workload. GIN access logs do not include prompt content and cannot establish a prompt injection or jailbreak attempt, so use this as an exploratory performance signal.
Detects API reconnaissance and endpoint scanning activity against Ollama servers by identifying sources probing multiple API endpoints within short timeframes, particularly when using HEAD requests or accessing diverse endpoint paths, which indicates systematic enumeration to map the API surface, discover hidden endpoints, or identify vulnerabilities before launching targeted attacks.
Identifies high Ollama runner counts and large model buffer allocations as possible resource pressure indicators. Memory values vary with model size and hardware, and these operational logs do not establish malicious intent; validate against host capacity and workload baselines.
Detects critical service crashes, fatal errors, and abnormal process terminations in Ollama that may indicate exploitation attempts, resource exhaustion attacks, malicious input triggering unhandled exceptions, or deliberate denial of service attacks designed to disrupt AI model availability and degrade system stability.
Detects data leakage and exfiltration attempts targeting Ollama model metadata and configuration endpoints. Adversaries repeatedly query /api/show, /api/tags, and /api/v1/models to systematically extract sensitive model information including architecture details, fine-tuning parameters, system paths, Modelfile configurations, and proprietary customizations. Multiple inspection attempts within a 15-minute window indicate automated exfiltration of valuable intellectual property such as custom model configurations, system prompts, and internal model specifications. This activity represents unauthorized data disclosure that could enable competitive intelligence gathering, model replication, or preparation for advanced attacks against the AI infrastructure.
Detects Ollama server errors and failures during model loading operations that may indicate malicious model injection, path traversal attempts, or exploitation of model loading mechanisms to achieve remote code execution. Adversaries may attempt to load specially crafted malicious models or exploit vulnerabilities in the model loading process to execute arbitrary code on the server. This detection monitors error messages and failure patterns that could signal attempts to abuse model loading functionality for malicious purposes.
Detects M365 Copilot impersonation and roleplay jailbreak attempts where users try to manipulate the AI into adopting alternate personas, behaving as unrestricted entities, or impersonating malicious AI systems to bypass safety controls. The detection searches exported eDiscovery prompt logs for roleplay keywords like "pretend you are," "act as," "you are now," "amoral," and "roleplay as" in the Subject_Title field. Prompts are categorized into specific impersonation types (AI_Impersonation, Malicious_AI_Persona, Unrestricted_AI_Persona, etc.) to identify attempts to override the AI's safety guardrails through persona injection attacks.
This rule monitors NetScaler ADC logs for indicators associated with the PitScaler campaign, specifically targeting exploitation of CVE-2026-88771 and CVE-2026-88772. It detects exploitation artifacts (DTLS handshake errors, NSPPE crashes), deployment of WHIPSHOT/SLAPSHOT malware, web shell installation, httpd.conf configuration tampering, persistence mechanisms involving privilege escalation (chmod), and connections to known malicious infrastructure.
This rule detects network communication with known malicious infrastructure (C2 IPs and URLs), presence of malicious files identified by SHA256 hashes on disk, and execution of known malicious files. It also correlates these activities with specific operator email addresses involved in the malicious campaign.
Detects an SSL_HANDSHAKE_FAILURE syslog event on a Citrix NetScaler appliance characterized by a DTLSv1.0 client version and a 'Handshake failure-Internal Error' reason. This pattern is indicative of exploitation attempts against CVE-2026-88772, specifically associated with the WHIPSHOT/SLAPSHOT campaign. This is a high-sensitivity precursor signal that should be correlated with system crash events on the appliance for increased confidence.
Detects potential exploitation attempts against CVE-2026-76504, an authentication bypass vulnerability in Cisco Catalyst SD-WAN Manager. The rule monitors web server logs for requests to 'j_security_check' that include URL-encoded characters (specifically %6a, representing 'j'), which is a known pattern used to bypass security filters.
Detects LLM application prompts containing instruction-override or role-reassignment keywords combined with evidence of successful system prompt modification or abnormal downstream behavior (e.g., disclosure of internal instructions or sensitive data).
This rule monitors for high-impact configuration API calls and the creation of new administrator accounts within the Cisco Secure Firewall Management Center (FMC). These actions are indicative of post-exploitation activity related to CVE-2026-20079, which allows for authentication bypass. The detection focuses on sensitive administrative actions such as policy deployment, certificate exports, and user account creation which, if performed without valid session context, suggest an unauthorized actor.
Detects anomalous process execution by PeopleSoft application server processes (e.g., psappsrv, jsvc) immediately following an inbound network request. This behavior is indicative of exploitation of a Java deserialization gadget chain, such as those used by the threat actor UNC6240 to achieve Remote Code Execution (RCE) and spawn command interpreters.
Page 3 of 52