avatar

Daniel Jeremiah

@daniyyell
United KingdomTrusted contributorCompletionist
8 followers4,600 downloads2,430 copies68 likes128,312 views

978 detections

OysterLoader stage 3 creates a scheduled task (every 13 min) running rundll32 with a DLL from AppData. Task names include COPYING3, VisualUpdater, AlphaSecurity, DetectorSpywareSecurity. Sekoia report.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
1025
This rule detects HTTP network connections to the Telegram bot API (api.telegram.org/bot). This pattern is often associated with infostealers exfiltrating data to attacker-controlled Telegram bots. The rule specifically looks for the 'NETWORK_HTTP' event type and a URL containing '/api.telegram.org/bot' (case-insensitive).
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
2038
This rule detects the execution of specific download utilities (curl, wget, aria2c, fetch, PowerShell, certutil, bitsadmin, mshta) attempting to download 'install.sh' from either 'clawd.bot' or 'openclaw.ai'. This activity is indicative of potential malware or unauthorized script downloads, possibly bypassing network security controls.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
6043
This rule detects PHP webshell patterns. The first rule, 'PHP_Webshell_GatedEval_GETaaaa_POST1', identifies webshells that use a specific GET parameter 'aaaa' and a POST parameter '1' in conjunction with an 'eval' function to execute code. The second rule, 'PHP_Webshell_ASS_Base64_Eval_DQo_Delimiters', looks for webshells that utilize 'base64_decode' and 'eval' functions, often with a distinctive 'DQo=' delimiter, and receive input from at least two common web input sources like $_COOKIE, $_GET, or $_POST, along with the string 'ASS'.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
0016
This YARA rule detects the Arkanix Stealer by identifying specific strings found within its code or related artifacts. It looks for a combination of unique user agent strings, domain names associated with its command and control infrastructure, API endpoints used for session creation or data exfiltration, a secret key, and a staging identifier. The rule requires at least one domain, one endpoint, and one of the distinctive user agent, key, or staging hints to trigger, aiming to reduce false positives by combining multiple indicators.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
1029
Detects exploitation attempts of CVE-2026-22769 by identifying PUT requests directed at the Tomcat Manager application deployment endpoint. This is used by UNC6201 to deploy malicious WAR files containing the SLAYSTYLE webshell.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
0022
Detects a combination of suspicious registry persistence under the Microsoft TIP AggregateResults key and PowerShell launched with ExecutionPolicy Unrestricted referencing LunarApplication. Based on publicly described artefacts and intended for tuning and validation.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
0021
This rule detects the creation or modification of specific files associated with the SHub v2.0 infostealer on macOS. It looks for a fake 'GoogleUpdate' binary within the '~/Library/Application Support/Google/' directory or a LaunchAgent plist file named 'com.google.keystone.agent.plist' in the 'LaunchAgents/' directory. These file paths are indicative of the SHub stealer's persistence mechanism, which impersonates legitimate Google update components.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
0032
Detects modification of the 'convert_hosts.sh' script on Dell RecoverPoint appliances. This technique is used by UNC6201 to establish persistence for the GRIMBOLT and BRICKSTORM backdoors.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
0033
This rule detects attempts to exploit Ivanti EPMM vulnerabilities CVE-2025-4427 and CVE-2025-4428. The first signature identifies requests to the '/api/v2/featureusage' endpoint containing patterns indicative of remote code execution (e.g., '${', 'Runtime', 'getRuntime', 'exec'). The second signature detects requests for a '.jsp' file named '403', which could indicate a backdoor or web shell deployment post-exploitation.
avatar
Daniel Jeremiah@daniyyell
avatar
Detections.ai Community
9 months ago
0028
Page 81 of 98