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 attempts to bypass AWS IMDSv2 security protections via the LiteLLM /config/pass_through_endpoint, where attackers inject the 'x-pass-X-aws-ec2-metadata-token' header to potentially strip authentication requirements.
Detects requests to a LiteLLM proxy that have been assigned the 'PROXY_ADMIN' role despite the absence of valid authentication credentials (Authorization headers, JWT tokens, or OAuth2 tokens). This pattern typically indicates that the LiteLLM instance is misconfigured, specifically that a 'master_key' is missing, causing the proxy to default to a permissive mode where every request is treated as having administrative privileges.
Detects forged Citibank payment confirmation/account statement PDFs used in Phantom Deal BPC fraud to falsely demonstrate wire transfer progress; requires combination of bank branding with fraud-specific SWIFT/routing/CTA artifacts
Detects instances where node, npm, or bun package installation processes modify the Claude AI settings.json file. This could indicate potential supply chain compromise or unauthorized configuration changes to the Claude application environment via malicious packages or scripts.
Detects instances where node, npm, or bun package installation processes modify the Claude AI settings.json file. This could indicate potential supply chain compromise or unauthorized configuration changes to the Claude application environment via malicious packages or scripts.
Detects processes attempting to access sensitive Linux system files (/etc/passwd or /etc/shadow) that are not typically associated with system administration, security, or identity management tasks. This behavior may indicate an attempt by an unauthorized user or process to gather credentials or system configuration information.
This rule detects potential Server-Side Request Forgery (SSRF) activity where web application or runtime processes attempt to access the Cloud Instance Metadata Service (IMDS) or container task metadata endpoints. By monitoring network connections and application logs, the rule filters out known legitimate metadata clients and identifies suspicious processes frequently associated with web-based vulnerabilities that are repeatedly querying sensitive metadata paths.
This rule detects potential Server-Side Request Forgery (SSRF) activity where web application or runtime processes attempt to access the Cloud Instance Metadata Service (IMDS) or container task metadata endpoints. By monitoring network connections and application logs, the rule filters out known legitimate metadata clients and identifies suspicious processes frequently associated with web-based vulnerabilities that are repeatedly querying sensitive metadata paths.
Detects text/report artifacts (markdown, JSON, plain text) produced by an autonomous AI-driven vulnerability research pipeline that decompiles binaries, traces cross-references, hypothesizes memory-safety flaws, and generates/debugs proof-of-concept exploits
Detects anomalous activity patterns consistent with automated firmware reverse engineering pipelines, specifically correlating the high-frequency execution of firmware analysis tools (e.g., binwalk, Ghidra) with the creation of vulnerability research knowledge base artifacts. This behavioral heuristic aims to identify potential AI-driven or automated zero-day discovery workflows by tracking tool usage density and output characteristics.
Detects anomalous activity patterns consistent with automated firmware reverse engineering pipelines, specifically correlating the high-frequency execution of firmware analysis tools (e.g., binwalk, Ghidra) with the creation of vulnerability research knowledge base artifacts. This behavioral heuristic aims to identify potential AI-driven or automated zero-day discovery workflows by tracking tool usage density and output characteristics.
Detects the initialization of containerized LLM-based gateway services (e.g., litellm, ollama) followed closely in time by high-volume network traffic containing suspicious web injection payloads (e.g., SQLi, XSS, SSRF). The rule correlates process execution events for container launches with network activity originating from the same device to identify potential unauthorized or unsupervised automated exploitation.
This rule detects unauthorized or suspicious automated scraping behavior by identifying periodic, non-interactive tasks that perform high-volume outbound network requests to multiple external hosts. It specifically filters for processes launched via scheduling mechanisms (Task Scheduler, cron) that exhibit indicators of anti-bot or proxy-rotation bypass techniques (such as usage of headless browser tools or proxy-related keywords in command lines). The rule is tuned to ignore legitimate system maintenance, update, and monitoring processes.
This rule detects unauthorized or suspicious automated scraping behavior by identifying periodic, non-interactive tasks that perform high-volume outbound network requests to multiple external hosts. It specifically filters for processes launched via scheduling mechanisms (Task Scheduler, cron) that exhibit indicators of anti-bot or proxy-rotation bypass techniques (such as usage of headless browser tools or proxy-related keywords in command lines). The rule is tuned to ignore legitimate system maintenance, update, and monitoring processes.
Detects the creation, modification, or renaming of PHP files within the 'wp-content/uploads' directory of a WordPress installation. This pattern is commonly indicative of an attacker uploading a web shell to maintain persistence or execute arbitrary code on a compromised web server.
This rule detects suspicious process execution originating from web server applications (e.g., php-fpm, apache2, nginx). It looks for common shell interpreters or command-line utilities (e.g., bash, curl, python) triggered by these web server processes when they contain command execution function strings in their command lines (e.g., system, shell_exec, passthru), which is a common indicator of web shell activity or post-exploitation command execution.
Detects execution of shell interpreters (tar, sh, bash) by the Parallels Desktop dispatcher service (prl_disp_service) or usage of compression flags containing specific markers associated with UCP/ACTOR privilege escalation patterns, specifically when executed with root privileges or by a root user.
Detects interactive root shell requests via 'sudo -i' that occur shortly after a failed Parallels appliance installation or tar extraction event containing specific error codes or configuration markers (UCP_UID=0). This pattern aligns with the exploitation chain for CVE-2026-90894, where an attacker leverages an extraction failure to escalate privileges.
Detects the use of the 'tar' utility with the '--use-compress-program' flag, where the compression program specified is not a standard system compression tool. This pattern is commonly used by attackers to execute arbitrary payloads when tar extracts malicious archives.
Detects a Parallels appliance tar archive containing a directory entry name with an embedded double-quote character followed by the --use-compress-program flag, characteristic of the ParaShells (CVE-2026-90894) tar argument-injection payload
Detects suspicious execution of the 'tar' utility spawned by 'prl_disp_service' (Parallels Desktop service) using potentially malicious command-line arguments. The rule specifically flags the use of '--use-compress-program' or the inclusion of double quotes in the command line, which are often used to achieve arbitrary code execution via shell injection in this context.
Page 99 of 340