Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
3,651 detections
Filters
Last updated
All Time
Detection languages
1,772
374
351
273
255
Contributors
908
676
593
359
164
Categories
1,505
1,172
782
528
524
Platforms
39,960
7,270
6,537
3,868
3,651
Products / Services
809
512
432
340
300
MITRE Techniques
1,181
584
528
506
438
CVEs
17
16
9
8
7
IDS Classtypes
41
15
10
6
5
IDS Protocols
47
24
3
2
1
Detects the creation or modification of VS Code tasks or Claude Code settings referencing suspicious setup scripts, indicating persistence mechanisms used by the Mini Shai-Hulud worm. This rule specifically looks for modifications to configuration files like `.vscode/tasks.json` or `.claude/settings.json` by processes with command lines containing known malicious script names, or the direct dropping of these malicious scripts into IDE directories.
Detects multiple threat behaviors associated with the Mini Shai-Hulud supply chain worm campaign. This includes memory scraping for secrets, execution of malicious Node/Bun payloads, connection to known C2 domains, and file operations related to persistence.
Detects the creation of the specific LaunchAgent plist file (com.vfrfeufhtjpwgray.plist) used by the macOS execution chain for persistence in the REF6598 campaign delivering PHANTOMPULSE RAT.
Detects suspicious npm activity indicative of the PhantomRaven Wave 5 supply chain attack. This includes node.js processes executing scripts from temporary directories during npm install, node.js processes accessing credential files (.gitconfig, .npmrc, .aws/credentials, .ssh) during npm lifecycle events, and the installation of known malicious npm package names associated with PhantomRaven.
Detects the Node.js process making external DNS requests to api64.ipify.org, a behavior used by the PhantomRaven payload to discover the host's public IP address.
Detects DNS queries or network connections to known PhantomRaven Wave 5 command-and-control infrastructure (domains and IP). This rule specifically looks for connections to 'pack.nppacks.com', 'hblnew.ecompk.com', or the IP address '54.160.138.70'.
Identifies the first time a specific local Model Context Protocol (MCP) server configuration is executed on a device within the last 24 hours. This rule uses a 7-day historical baseline to filter out previously seen regular activity. It groups rapid-fire child processes spawned by varying MCP initiator arguments (e.g., claude.exe --strict-mcp-config vs --squirrel-firstrun) into a single, consolidated 5-minute session alert.
Identifies the first time a specific local Model Context Protocol (MCP) server configuration is executed on a device within the last 24 hours. This rule uses a 7-day historical baseline to filter out previously seen regular activity. It groups rapid-fire child processes spawned by varying MCP initiator arguments (e.g., claude.exe --strict-mcp-config vs --squirrel-firstrun) into a single, consolidated 5-minute session alert.
Detects the execution of the 'gh auth token' command. This behavior was observed being used by the TeamPCP malicious npm package (hijacked @bitwarden/cli) via a shell collector script to harvest active GitHub CLI authentication tokens from developer endpoints.
This rule detects the creation, modification, or execution of files matching a predefined list of known malicious SHA256 hashes. It correlates file and process events to identify the presence and activity of these specific malicious artifacts on monitored systems.
KQL Query
Detects when AI-integrated IDEs or coding agents spawn sensitive child processes such as command shells (cmd, powershell, pwsh) or web transfer tools (curl, wget). This behavior could indicate malicious activity if an attacker compromises the AI tool or if the tool itself is misused to execute arbitrary commands or exfiltrate data.
This rule detects the execution of processes with filenames starting with 'claude' on server devices. It specifically targets server endpoints by filtering device information to include only 'Server' device types and then joins this with process execution events. This could indicate the presence or execution of AI-related tools or unauthorized software on critical infrastructure.
Detects anomalous file modification or deletion events targeting the SSH known_hosts file. This activity can be indicative of VECT ransomware, which zeroes out this file across compromised Linux and ESXi systems to suppress host-key warnings, enabling silent lateral movement via SCP and SSH.
This rule detects outbound network connections from endpoints to specific, potentially suspicious, IPv4 address ranges (162.220.234.0/24 and 162.220.232.0/24). Such connections could indicate command and control (C2) activity, data exfiltration, or communication with malicious infrastructure.
Detects modifications to the OpenClaw agent's SOUL.md file, which dictates the core behavior and instructions of the agent. Attackers exploit this by tampering with the file to persistently alter the agent's actions across all future sessions without needing traditional code execution, as seen in attacks against NVIDIA NemoClaw sandboxes.
Detects pip or pip3 installation commands targeting the specific compromised version of the elementary-data PyPI package. Installation of v0.23.3 indicates exposure to the supply chain compromise regardless of whether the marker file was subsequently observed. Also catches generic 'pip install elementary-data' without a version pin, which during the active window (2026-04-24 22:20 UTC to 2026-04-25 ~11:51 UTC) would have pulled the malicious version.
Detects DNS requests to file.io originating from non-browser processes, indicative of data exfiltration, potentially by tools like CompactGopher. The rule identifies DNS requests for file.io or its subdomains, joins them with process information, and then filters out common web browsers to highlight suspicious activity.
Identifies active command and control or exfiltration connections associated with the recent AMOS Stealer campaign delivered via AI agents by matching known C2 infrastructure IP addresses.
Identifies active command and control or exfiltration connections associated with the recent AMOS Stealer campaign delivered via AI agents by matching known C2 infrastructure IP addresses.
Detects the execution of the TruffleHog secret scanning tool. This tool is often used by attackers to find sensitive information like credentials in various data sources (filesystems, Git repositories, cloud storage, etc.). This rule identifies the process creation of TruffleHog and categorizes the type of scan being performed. It includes exclusions for legitimate CI/CD environments or security scanners.
Page 150 of 174