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
Detects abnormal network activity and connectivity issues in Ollama including non-localhost API access attempts and warning-level network errors such as DNS lookup failures, TCP connection issues, or host resolution problems that may indicate network-based attacks, unauthorized access attempts, or infrastructure reconnaissance activity.
Detects potential prompt injection or jailbreak attempts against Ollama API endpoints by identifying requests with abnormally long response times. Attackers often craft complex, layered prompts designed to bypass AI safety controls, which typically result in extended processing times as the model attempts to parse and respond to these malicious inputs. This detection monitors /api/generate and /api/chat endpoints for requests exceeding 30 seconds, which may indicate sophisticated jailbreak techniques, multi-stage prompt injections, or attempts to extract sensitive information from the model.
Detects potential Distributed Denial of Service (DDoS) attacks or rate limit abuse against Ollama API endpoints by identifying excessive request volumes from individual client IP addresses. This detection monitors GIN-formatted Ollama server logs to identify clients generating abnormally high request rates within short time windows, which may indicate automated attacks, botnet activity, or resource exhaustion attempts targeting local AI model infrastructure.
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.
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.
The following analytic detects the creation of files in the Windows %startup% folder, a common persistence technique. It leverages the Endpoint.Filesystem data model to identify file creation events in this specific directory. This activity is significant because adversaries often use the startup folder to ensure their malicious code executes automatically upon system boot or user logon. If confirmed malicious, this could allow attackers to maintain persistence on the host, potentially leading to further system compromise and unauthorized access to sensitive information.
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.
The following analytic identifies the execution of RAR utilities to archive files on a system. It leverages data from Endpoint Detection and Response (EDR) agents, focusing on process names, GUIDs, and command-line arguments. This activity is significant as threat actors, including red-teamers and malware like DarkGate, use RAR archiving to compress and exfiltrate collected data from compromised hosts. If confirmed malicious, this behavior could lead to the unauthorized transfer of sensitive information to command and control servers, posing a severe risk to data confidentiality and integrity.
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. The rule aggregates multiple termination events within a 5-minute window to identify sustained instability or attack patterns.
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 abnormal memory allocation patterns and excessive runner operations in Ollama that may indicate resource exhaustion attacks, memory abuse through malicious model loading, or attempts to degrade system performance by overwhelming GPU/CPU resources. Adversaries may deliberately load multiple large models, trigger repeated model initialization cycles, or exploit memory allocation mechanisms to exhaust available system resources, causing denial of service conditions or degrading performance for legitimate users.
Detects changes to payment or bank account information in Workday. This activity is associated with 'payroll pirate' attacks by threat actor Storm-2657, who modify payroll details to divert salary payments after compromising an account.
KQL Query
Detects a series of Salesforce Object Query Language (SOQL) queries indicative of data reconnaissance and exfiltration, similar to techniques used by UNC6395.
The following analytic identifies attempts to access the Remote ShellServlet on a web server, specifically targeting Confluence servers vulnerable to CVE-2023-22518 and CVE-2023-22515. It leverages web data to detect URLs containing "*plugins/servlet/com.jsos.shell/*" with a status code of 200. This activity is significant as it is commonly associated with web shells and other malicious behaviors, potentially leading to unauthorized command execution. If confirmed malicious, attackers could gain remote code execution capabilities, compromising the server and potentially the entire network.
Detects processes potentially related to XillenStealer making outbound network connections or DNS queries for telegram.org. This is indicative of command and control or data exfiltration activity.
This rule detects the potentially malicious installation of VMware Installation Bundles (VIBs) on an ESXi host using the `--force` flag. The `--force` option bypasses critical signature and compatibility checks, which is uncommon in normal administrative operations. This behavior is often observed in post-compromise scenarios where adversaries install backdoored or unauthorized kernel modules, drivers, or monitoring tools to establish persistence or gain deeper control of the hypervisor.
This analytic detects the configuration of TFTP services on Cisco IOS devices that could be used to exfiltrate sensitive configuration files. Threat actors like Static Tundra have been observed configuring TFTP servers to make device configuration files accessible for exfiltration after gaining initial access. The detection specifically looks for commands that expose critical configuration files such as startup-config, running-config, and other sensitive system information through TFTP. This activity is particularly concerning as it may represent an attempt to steal credentials, network topology information, and other sensitive data stored in device configurations.
Detects potential exploitation of a critical ABAP code injection vulnerability (CVE-2025-42957) in SAP S/4HANA. An attacker with low-level privileges can leverage a vulnerable RFC module, /SLOAE/DEPLOY, to execute arbitrary code and gain full control of the system.
Detects Salesforce Object Query Language (SOQL) queries that may be attempting to access or exfiltrate sensitive credentials like passwords, API keys, or tokens stored within Salesforce objects. This activity was observed in the UNC6395 campaign where attackers targeted AWS keys, Snowflake tokens, and passwords.
Detects data theft activity from Salesforce instances by the threat actor UNC6395, who leveraged compromised OAuth tokens associated with the Salesloft Drift third-party application. The detection is based on suspicious user agents and known malicious IP addresses used by the threat actor, specifically targeting the 'Drift' connected application.
Page 47 of 52