Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
2,982 detections
Filters
Last updated
All Time
Detection languages
1,486
513
403
239
167
Contributors
986
658
304
198
109
Categories
916
822
673
590
452
Platforms
39,932
7,263
6,535
3,865
3,643
2,982
Products / Services
1,229
703
443
309
200
MITRE Techniques
1,053
824
459
384
348
CVEs
17
16
16
10
10
IDS Classtypes
24
7
7
4
3
IDS Protocols
49
1
Detects successful authentication to administrative identity provider consoles that are flagged by identity-provider risk engines as anomalous (e.g., unfamiliar location, new device, or suspicious sign-in properties). This pattern monitors for potential account takeovers or compromised insider activity targeting critical administrative infrastructure, as observed in attacks utilizing recruited or purchased credentials.
Detects high-volume data exports or bulk API queries against sensitive Salesforce CRM objects (Accounts, Contacts, Cases, Opportunities) performed by connected applications. This pattern is characteristic of exfiltration efforts where a newly authorized OAuth application is used to quickly extract large quantities of sensitive records.
Detects anomalous engineering account activity associated with supply-chain compromise, specifically identifying the generation of access tokens or retrieval of secrets followed by unauthorized changes to CI/CD pipeline configurations or secret stores in platforms like GitHub, JFrog Artifactory, or BrowserStack.
Detects OAuth application consent requests that combine broad API access scopes with persistence-enabling scopes (e.g., offline_access or refresh_token). This pattern is consistent with OAuth consent phishing attacks, where adversaries attempt to gain non-interactive, long-term access to an organization's cloud environment via malicious connected applications.
Detects unauthorized attempts to delete or purge critical Salesforce audit logs, including the Setup Audit Trail, Login History, and Event Monitoring log files. This behavior is indicative of anti-forensic activity, often utilized by threat actors such as ShinyHunters to destroy audit evidence following unauthorized access, data export, or credential manipulation.
Detects high-volume, individually-tailored spear-phishing campaigns likely assisted by Generative AI. The rule monitors for a single sender targeting a high number of unique recipients with distinct, non-identical email subjects within a short timeframe. It specifically flags activity involving newly registered look-alike domains and interactions with credential-harvesting patterns in URLs.
This rule detects potential prompt injection attacks against internal AI-powered assistants, agents, or LLM gateways. It monitors network HTTP requests for common patterns used to override system instructions, bypass safety filters, or force the model to adopt a privileged, unauthorized persona (e.g., administrator or debug mode). These techniques are often used in social engineering to manipulate the AI's output or security posture.
Detects HTTP requests to popular LLM API services (OpenAI, Anthropic, Google, Cohere) containing common jailbreak or role-play patterns. These patterns attempt to bypass safety guardrails by using techniques such as persona-switching (e.g., 'DAN'), hypothetical scenario framing, or unauthorized security testing pretexts.
Detects modifications to Entra ID Conditional Access or Okta authentication policies that introduce MFA exclusions, trusted network bypasses, or authentication-strength downgrades. This activity is highly indicative of an adversary attempting to maintain persistence or bypass MFA following a successful initial account compromise.
Detects suspicious account management activities, such as password resets or MFA re-enrollment, performed by an operator on a high-privileged or administrator account outside of typical business hours (06:00-20:00). This behavior is often indicative of help-desk impersonation or credential takeover attempts targeting Identity Provider (IdP) administrative roles.
Detects a suspicious sequence of events indicating account takeover: a password reset, followed by an MFA factor re-enrollment, and concluding with a successful login from a geographically impossible location compared to recent history. This pattern often signifies a help-desk-assisted account compromise or token theft.
Detects sign-in activity using the OAuth 2.0 device code flow, followed by subsequent authentication attempts from the same account originating from anomalous, previously unseen geographic locations or using unfamiliar applications. This pattern is indicative of a device code phishing attack where an adversary lures a user into authorizing a malicious device or application and subsequently hijacks the resulting session to access resources from a different location.
This rule detects anomalous, high-volume activity in Jira, such as mass ticket viewing, searching, exporting, or attachment downloading, which is often characteristic of unauthorized data exfiltration or reconnaissance. The detection triggers when an account exceeds thresholds of 200 issue-related actions or 8 attachment downloads within a 7-day period.
Detects anomalous, high-volume querying against machine learning inference endpoints. The rule identifies potential model extraction or data exfiltration attempts by flagging high counts of requests from a single user/API key, particularly during off-hours or with high variance in unique inputs, which may indicate automated efforts to reconstruct proprietary models or extract sensitive training data.
Detects anomalous, high-volume querying against machine learning inference endpoints. The rule identifies potential model extraction or data exfiltration attempts by flagging high counts of requests from a single user/API key, particularly during off-hours or with high variance in unique inputs, which may indicate automated efforts to reconstruct proprietary models or extract sensitive training data.
Detects potential data leakage from an LLM application where a user repeatedly prompts the model to divulge its internal system instructions, configuration, or sensitive information (e.g., API keys, PII) and the model subsequently outputs that sensitive data.
Detects attempts to perform context or memory poisoning against an AI agent by injecting imperative, instruction-like payloads into the agent's long-term memory or context store. The rule specifically monitors updates originating from untrusted channels, such as tool outputs, external documents, or web content, which contain patterns commonly associated with prompt injection attacks (e.g., overriding system instructions or role definitions).
Detects an AI agent session that performs a data exfiltration action (e.g., sending an email, webhook, or file share) to an unauthorized or non-allowlisted destination shortly after the agent has ingested potentially untrusted external content. This behavior is indicative of an AI agent being manipulated to exfiltrate data after processing malicious or adversarial inputs.
Detects the ingestion of sensitive credentials (API keys, passwords, private keys) into a RAG (Retrieval-Augmented Generation) document store, followed by the retrieval of those credentials by an application agent, indicating potential credential harvesting within the LLM application pipeline.
Detects unauthorized modifications to AI Agent Model Context Protocol (MCP) tool manifests. This includes the injection of imperative instructions into tool descriptions, the addition of suspicious instruction-related fields to tool output schemas, or manifest updates initiated by unverified publishers.
Detects the installation of Model Context Protocol (MCP) servers or AI agent plugins that exhibit characteristics of supply chain risk, such as unverified or unknown publishers, requests for highly sensitive permissions (filesystem write, credential read, unrestricted network egress), or recent public listing, which aligns with AI model supply chain poisoning techniques (AML.T0104).
Page 30 of 142