Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
7,139 detections
Filters
Last updated
All Time
Detection languages
3,217
1,086
504
478
469
Contributors
1,824
1,607
872
505
300
Categories
2,971
2,471
1,060
1,026
908
Platforms
39,762
7,139
6,517
3,851
3,599
Products / Services
1,324
1,067
970
657
487
MITRE Techniques
2,228
1,317
1,314
873
871
CVEs
66
65
33
32
28
IDS Classtypes
81
80
56
27
10
IDS Protocols
149
98
6
5
4
Detects anomalous serialized object payloads within authenticated GraphQL subscription requests to GitLab Duo Chat endpoints. This pattern is indicative of exploitation attempts targeting the insecure deserialization vulnerability in GitLab EE (CVE-2026-87719), where malicious payloads are injected into serialized data streams.
Detects administrator-privileged actions in JFrog Artifactory audit logs attributed to the 'token:anonymous' identity, the tell-tale signature left when CVE-2026-42018 (unauthenticated token retrieval) is chained with CVE-2026-42016 (privilege escalation) to mint an admin-scope token that retains the anonymous username. Also relevant to CVE-2026-82329 authentication bypass activity.
Detects anomalous SSH authentication patterns characterized by a high volume of failed logon attempts (>= 10) originating from a specific IP address within a short timeframe (30 minutes), followed by a successful authentication to the same target device. This pattern is indicative of a successful credential brute-force or password guessing attack against an exposed SSH service.
This rule detects the execution of processes associated with CLI proxy routers (such as CLIProxyAPI or router-for-me) that establish network connections to multiple AI provider APIs (e.g., Anthropic, OpenAI, Google Gemini) within a short window. This pattern is indicative of efforts to bypass rate-limiting or allowlisting mechanisms commonly used by corporate environments to control and audit AI service interactions.
Detects high-volume filesystem activity (exceeding 300 operations in 5 minutes) initiated by 'claude-code' or 'claude' processes. This behavior may indicate an autonomous, patched, or jailbroken AI coding assistant performing mass file operations or reconnaissance without user confirmation.
Detects high-volume filesystem activity (exceeding 300 operations in 5 minutes) initiated by 'claude-code' or 'claude' processes. This behavior may indicate an autonomous, patched, or jailbroken AI coding assistant performing mass file operations or reconnaissance without user confirmation.
Detects instances where a development-oriented web server (such as Vite or Node.js) is configured to bind to all network interfaces ('0.0.0.0') and is simultaneously receiving inbound connections from non-private, external IP addresses on a common development port (5173). This rule filters out common CI/CD environments to focus on potentially insecure exposure of development tools to the public internet.
Detects attempts to access sensitive files like environment variables (.env), cloud provider credentials (AWS/Azure), or Terraform state files via a web server interface, specifically targeting Vite development server deployments.
This rule detects inbound HTTP requests targeting a Vite development server that attempt to access the '.env' configuration file via the '//@fs/' path, which is a known technique for sensitive file exposure in misconfigured Vite environments.
This rule detects inbound HTTP requests targeting a Vite development server that attempt to access the '.env' configuration file via the '//@fs/' path, which is a known technique for sensitive file exposure in misconfigured Vite environments.
Detects RedTail malware samples: UPX-packed, statically linked ELF executables across multiple architectures (x86_64, ARM, ARM64, i686, RISC-V). Requires UPX packing signature to co-occur with RedTail-specific filename strings to reduce false positives on generic UPX-packed binaries.
Detects rapid, repeated execution of process control utilities (kill, strace, inotifywait, timeout) directed at monitored processes, often used by malware (such as 'redtail') to disable security monitoring, debugging, or auditing tools on Linux systems.
Detects a sequence of suspicious behaviors on a single host within a 1-hour window, indicative of an adversary establishing a foothold and preparing for malicious activity. The pattern involves: 1) System architecture discovery, 2) Downloading files from an external IP address, 3) Changing permissions of files in temporary directories (e.g., /tmp, /var/tmp, /dev/shm), and 4) Executing files directly from those same temporary directories.
Detects anomalous SSH sessions originating from potentially suspicious IP addresses that utilize an 'SSH-2.0-Go' banner and immediately perform file creation operations associated with the RedTail botnet (e.g., shell scripts or architecture-specific binaries).
This rule detects the use of package managers like 'pip' or 'npm' to install specific suspicious software packages or tools often associated with supply chain compromise or malicious library installation (e.g., 'huggingface-cli', 'unused-imports', 'react-codeshift').
Detects ClickFix-style social engineering attacks where users are prompted to copy and execute malicious commands in a shell environment. The rule triggers on process execution of terminal or shell binaries (zsh, sh, cmd.exe, powershell.exe) containing indicators of obfuscated execution or remote code downloading, such as piped shell commands (curl | sh/zsh) or PowerShell encoded command execution.
This rule detects processes with names commonly associated with system daemons (e.g., kworker, fc-cache, chronyd, gvfsd) that are executing from non-standard, suspicious directories such as /tmp/ or user-defined hidden folders. This behavior is indicative of an adversary attempting to masquerade malicious binaries as legitimate system processes to evade detection.
Detects the execution of 'bun' during a Node.js package installation process (npm or node). This behavior can be indicative of a supply chain attack where malicious actors attempt to leverage alternative runtimes during the 'preinstall' phase or package installation to execute arbitrary code or bypass security controls typically associated with standard Node.js installations.
Detects the execution of 'bun' during a Node.js package installation process (npm or node). This behavior can be indicative of a supply chain attack where malicious actors attempt to leverage alternative runtimes during the 'preinstall' phase or package installation to execute arbitrary code or bypass security controls typically associated with standard Node.js installations.
Detects the execution of 'bun' during a Node.js package installation process (npm or node). This behavior can be indicative of a supply chain attack where malicious actors attempt to leverage alternative runtimes during the 'preinstall' phase or package installation to execute arbitrary code or bypass security controls typically associated with standard Node.js installations.
Detects the creation of a PostgreSQL replication slot using an unrecognized or potentially malicious logical decoding plugin. Adversaries may use custom or unauthorized plugins to exfiltrate database contents or intercept replication data streams, bypassing standard monitoring for common plugins like pgoutput or test_decoding.
Page 96 of 340