Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
1,733 detections
Filters
Last updated
All Time
Detection languages
909
387
164
127
68
Contributors
479
417
201
177
61
Categories
20,440
11,814
5,871
5,145
4,864
1,733
Platforms
822
436
329
274
206
Products / Services
498
378
118
96
79
MITRE Techniques
433
327
225
196
196
CVEs
12
12
10
10
6
IDS Classtypes
1
1
IDS Protocols
2
This search surfaces AWS Bedrock Claude prompts containing phrases commonly associated with prompt injection or jailbreak attempts, such as instruction overrides, persona switching, or requests to ignore prior guidance. These strings are not inherently malicious on their own - many of them (for example "act as", "you must", or "new persona") appear routinely in legitimate system prompts, few-shot examples, and role-play or creative-writing use cases. This is a hunting search rather than an anomaly detection because the base rate of benign matches is high; use it to pivot into the surrounding conversation and look for corroborating signals such as the phrase appearing mid-conversation rather than in an expected system/few-shot prompt, the user attempting to override guardrails or safety instructions rather than set a benign persona, instructions that follow the injected phrase requesting privilege escalation, data exfiltration, or access to restricted tools/resources, repeated or iterative attempts from the same user/role after being refused, and injected phrases originating from untrusted input (e.g., retrieved documents or tool output) rather than the human user turn.
This detection identifies potential cross-region inference abuse in AWS Bedrock Claude models. Cross-region inference abuse may indicate attempts to bypass regional restrictions, exfiltrate data, or perform unauthorized actions across different AWS regions.
Detects crates.io publish events combining an impersonation indicator (attacker account names/forged author email) with one of the known malicious package names, indicating a compromised or impersonated maintainer account was used to publish the Rust supply-chain payload.
Detects anomalous human-like interactive browsing behavior from the 'qa@snowflake.net' Jira service account. This account is typically restricted to automated ticket creation via GitHub Actions; broad exploration of Jira boards, timelines, and reports indicates a potential account compromise where an API token has been exfiltrated and repurposed for reconnaissance.
This rule detects browser-based runtime evaluations that utilize WebRTC's RTCPeerConnection and createDataChannel APIs, often indicative of an adversary attempting to establish a covert command-and-control (C2) channel directly from a compromised browser session.
Aggregates identity, endpoint, cloud application, and messaging telemetry to build a consolidated profile of the devices, clients, applications, and user agents associated with a specific user account.
The query combines Microsoft Entra sign-in data, endpoint logon activity, cloud application events, and email telemetry to provide a high-level overview of a user's activity footprint across the environment.
This query is useful for incident response, account compromise investigations, insider threat investigations, and validating whether a user is accessing services from expected devices and client applications.
The query combines Microsoft Entra sign-in data, endpoint logon activity, cloud application events, and email telemetry to provide a high-level overview of a user's activity footprint across the environment.
This query is useful for incident response, account compromise investigations, insider threat investigations, and validating whether a user is accessing services from expected devices and client applications.
Correlates Microsoft Entra Privileged Identity Management (PIM) role eligibility with actual activation activity to identify how privileged roles are being used across the environment.
The query combines eligible PIM assignments with activation events and submitted business justifications, allowing security teams to determine which privileged roles are frequently activated, rarely used, or never activated.
This is useful for privileged access reviews, least-privilege initiatives, role cleanup exercises, and validating whether users are providing meaningful activation justifications.
The query combines eligible PIM assignments with activation events and submitted business justifications, allowing security teams to determine which privileged roles are frequently activated, rarely used, or never activated.
This is useful for privileged access reviews, least-privilege initiatives, role cleanup exercises, and validating whether users are providing meaningful activation justifications.
Detects when a non-system application on an Android device is granted Accessibility Service permissions or Notification Listener access. Malicious actors frequently abuse these permissions to perform overlay attacks, keylogging, or intercepting sensitive information such as 2FA codes.
This rule detects the creation of a new WordPress administrator account within a short window (15 minutes) of an anomalous authentication or registration event on the same account or session. It is designed to identify post-exploitation activity, specifically persistence mechanisms following the exploitation of web application vulnerabilities like arbitrary file uploads or authentication bypasses in WordPress plugins.
This detection identifies prompts with hostile or aggressive sentiment being sent to AWS Bedrock Claude models. Hostile prompts may indicate potential abuse, harassment, or other malicious intent.
Detects user activity involving unsanctioned or 'Shadow IT' applications that have a low or poor Cloud Confidence Index (CCI) rating, as reported by Netskope. This rule helps identify potential security risks associated with the use of unapproved or newly observed cloud services.
Detects when a user authorizes an OAuth application that requests high or critical permissions in a SaaS environment via Netskope logs. This pattern is commonly used in OAuth-based phishing campaigns or illicit application consent grants where adversaries attempt to gain persistent access to cloud resources.
Detects unauthorized macOS Screen Sharing (screensharingd) attachment attempts using the 'root' user account via the Secure Remote Password (SRP) authentication mechanism, specifically targeting an authentication bypass vulnerability.
Detects instances where Zscaler AI Protect (AI Guard) flags attempts to perform prompt injection or jailbreak against Large Language Model (LLM) or GenAI applications. The rule monitors blocked requests and allowed requests that exceed a high confidence score threshold (0.75), identifying potential malicious interactions with generative AI services.
KQL Query
Detects high-risk user behavior associated with unsanctioned or restricted Generative AI applications, specifically targeting large data uploads (potential data exfiltration) or repeated access following a coaching intervention.
Detects DLP policy violations flagged by Netskope when users upload, share, post, or send files within Slack or Microsoft Teams. This rule monitors for DLP profile matches which may indicate unauthorized exfiltration or exposure of sensitive data via collaboration tools.
This detection identifies sensitive data such as social security numbers, passwords, API keys, and credit card numbers being sent in prompts to AWS Bedrock Claude models. Exposure of sensitive data through AI prompts may indicate data loss, credential leakage, or insider threat activity.
This detection identifies unusually large prompts sent to AWS Bedrock Claude models by calculating the statistical baseline of input token counts and flagging requests that exceed one standard deviations above the mean. Abnormally large prompts may indicate prompt injection attacks, data exfiltration attempts, or abuse of the AI service.
FortiGate High CPU Utilization Detected
Cortex XDR
Detects FortiGate devices whose average reported CPU utilization exceeds 90 percent during the correlation window. Sustained CPU pressure can degrade firewall throughput, session establishment, inspection performance, and management responsiveness and should be investigated as a device-health condition.
Detects instances where Zscaler AI Protect (AI Guard) flags attempts to perform prompt injection or jailbreak against Large Language Model (LLM) or GenAI applications. The rule monitors blocked requests and allowed requests that exceed a high confidence score threshold (0.75), identifying potential malicious interactions with generative AI services.
Page 49 of 83