avatar

Tim Peck

@timpeck
Trusted contributorCompletionist
31 followers5,374 downloads4,081 copies490 likes318,454 views

919 detections

Detects Meow RAT (ClickFix campaign) copying Chromium browser extension data directories associated with cryptocurrency wallets, including MetaMask (nkbihfbeogaeaoehlefnkodbefgpgknn), via osascript on macOS.
avatar
Tim Peck@timpeck
avatar
Detections.ai Community
5 months ago
0024
Detects instances where 'MpClient.dll', a legitimate Microsoft Defender component, is loaded by a process that is not a known, trusted Microsoft Defender executable. This behavior can indicate an attempt to abuse Defender's scanning or cleaning APIs for malicious purposes, as seen in 'RoguePlanet' style attacks.
avatar
Tim Peck@timpeck
avatar
Detections.ai Community
5 months ago
962399
Detects persistence from the postcss-minify-selector-parser npm typosquat Windows RAT (JFrog, Jun 2026). The Python/Nuitka implant (launched as chost.exe, a renamed python.exe running loader.py) writes an HKCU Run value named 'csshost'. Fires on a CurrentVersion\Run write where the value name is csshost OR the writing process is chost.exe.
avatar
Tim Peck@timpeck
avatar
SecOps Signals
5 months ago
2017
Detects enabling of Developer Mode via AllowDevelopmentWithoutDevLicense under AppModelUnlock. This is the sole precondition for the UNCanny technique (loose-file AppX registration from a UNC path -> SYSTEM LoadLibraryW coercion/LPE via the Windows Store InstallService). Low fidelity in isolation - Developer Mode is common on dev endpoints; correlate with UNC loose-file registration and svchost UNC image loads.
avatar
Tim Peck@timpeck
avatar
Detections.ai Community
5 months ago
1046
Detects post-exploitation of CVE-2026-22778 (vLLM libopenjp2 JPEG2000 cdef heap overflow). The overwritten AVBuffer free() pointer is set to system(), so a successful exploit spawns /bin/sh -c as a direct child of the vLLM/python inference process. Matches a shell or download/exec utility launched with a vLLM/python/uvicorn parent, excluding benign python worker children.
avatar
Tim Peck@timpeck
avatar
SecOps Signals
5 months ago
0056
Detects Add-AppxPackage -Register against a manifest on a UNC path. Legitimate loose-file registration during development points at local paths; a UNC InstalledLocation is the core of UNCanny - it hands the SYSTEM Windows Store InstallService a LoadLibraryW on an attacker-controlled UNC via plugin resolution.
avatar
Tim Peck@timpeck
avatar
Detections.ai Community
5 months ago
2029
Detects the ClickFix social engineering technique where a user is coerced into copying and executing a malicious script from a web browser. The rule identifies instances where common web browsers spawn script interpreters such as powershell.exe, cmd.exe, mshta.exe, or wscript.exe, which is characteristic of the user pasting malicious commands into a command prompt interface.
avatar
Tim Peck@timpeck
avatar
Detections.ai Community
5 months ago
005
Detects attempts to create rogue WordPress administrator accounts using indicators associated with the OptinMonster/TrustPulse supply chain attack. The rule specifically looks for POST requests to WordPress user creation endpoints (REST API, admin form, AJAX) containing campaign-specific identifiers like 'developer_api1', 'customer1usx', or patterns like 'dev_xxxxxx' in the response body or request URL. This covers the known vectors of the attack, aiming to identify the unauthorized creation of administrative users.
avatar
Tim Peck@timpeck
avatar
SecOps Signals
5 months ago
3036
Detects the command line sequence used by Miasma to restore passwordless sudo for the 'runner' user in GitHub Actions by overwriting sudoers via a Docker container. This specific pattern targets the Miasma Alpine container payload.
avatar
Tim Peck@timpeck
avatar
Detections.ai Community
5 months ago
6062
Detects the creation of a new deviceManagementScript or deviceShellScript object in Intune, indicating a PowerShell (Windows) or Shell (macOS/Linux) script has been uploaded to the Intune management platform. This is considered a high-fidelity indicator of script-based abuse as these scripts execute as SYSTEM on targeted endpoints and leave minimal forensic traces. The rule also identifies script creation initiated by service principals that might bypass human-UPN-based monitoring.
avatar
Tim Peck@timpeck
avatar
Detections.ai Community
5 months ago
220109
Page 7 of 92