Driver intelligence
LOLDrivers
Research vulnerable and malicious Windows drivers, inspect known samples, and understand how attackers use them.
Driver samples and abuse context
Explore LOLDrivers (opens in a new tab)MagicSword Intelligence
Attackers often skip custom malware and use what is already trusted: built-in utilities, signed drivers, remote access tools, and code-signing certificates. MagicSword brings public research and our own curated lists together so you can decide what runs on Windows, macOS, and Linux.
Free forever. No credit card required.
From research to decision
Here is how one listed tool moves from public research to a decision on your endpoints.
Illustrative example · AnyDesk
01 · Research
LOLRMM documents AnyDesk: where it installs, what it leaves behind, and how intruders have used it for remote access.
02 · Context
An endpoint launches AnyDesk.exe and MagicSword links the event to that LOLRMM entry. The match explains why the tool deserves a look. It does not make the file malicious; your help desk may rely on it.
03 · Decision
You might allow AnyDesk on support devices and block it on servers that should never run remote access tools. A tool is blocked only when a policy you configure says so.
Sources
A selection of the public catalogs and software publishers behind MagicSword intelligence, grouped by what they describe. Each link goes to the source itself.
Find a research source by name, platform, or topic.
Operating system binaries and scripts that attackers repurpose because they are already present and trusted.
Maintained by the LOLBAS Project
Windows binaries, scripts, and libraries with documented abuse such as download, execution, and defense bypass.
Maintained by the LOOBins community
Built-in macOS binaries and the ways attackers have used them.
Maintained by the GTFOBins community
Unix binaries that can be misused to escape restrictions, escalate privileges, or transfer files.
Real products with valid uses that also help intruders. A listing is not a verdict; it is a prompt to decide where the software belongs.
MagicSword community project
Remote monitoring and management tools, the artifacts they leave, and how attackers use legitimate remote access.
Maintained by mthcht
Tools threat actors have used to move data out of an environment.
Published by Microsoft
Administration and diagnostic utilities that are as useful to an intruder as they are to IT.
Published by NirSoft
Freeware utilities, including password-recovery tools that can expose stored credentials.
Published by joeware
Active Directory and Windows administration utilities, such as AdFind, that can support reconnaissance.
Components that load with elevated trust in the kernel or the browser.
MagicSword community project
Vulnerable and malicious drivers used to reach the kernel or disable security tools.
Maintained by mthcht
Malicious, compromised, and suspicious browser extensions.
Certificates that make malicious files look trustworthy at a glance.
Maintained by Squiblydoo
Abused code-signing certificates, with a focus on certificates sold to criminal operators.
Operated by abuse.ch
A code-signing certificate blocklist drawn from malware samples shared with the community.
Our research team maintains these curated lists to help you turn emerging threats into application control decisions.
Security-Tool Tampering
Tools built to disable, blind, or tamper with endpoint security products.
Exploited Code-Signing Certificates
Certificates seen signing malware in active campaigns, including compromised, expired, and revoked ones.
Windows and macOS Blocklists
Drivers, binaries, and certificates tied to known exploitation, plus macOS abuse paths such as user-writable locations, launchd, and script hosts.
Offensive and Dual-Use Tools
Security tooling that standard workstations rarely need. A file name alone is not proof of compromise, and approved security teams may need exceptions.
AI Coding Agents
Agentic tools that can run code, open shells, and read files, for teams that want to govern where they run.
Which sources are available depends on your plan. Compare plans
Community Projects
Public research and practical tools from the MagicSword team. Explore vulnerable drivers, remote access, agent skills, detection rules, and ways to test your defenses.
Driver intelligence
Research vulnerable and malicious Windows drivers, inspect known samples, and understand how attackers use them.
Driver samples and abuse context
Explore LOLDrivers (opens in a new tab)Remote access intelligence
Identify remote monitoring and management tools, their artifacts, and the ways threat actors can abuse legitimate remote access.
RMM tools and detection artifacts
Explore LOLRMM (opens in a new tab)Agent skill security
Inspect agent skill instructions, scripts, and bundled files for risky behavior before installation. Review evidence for the version you plan to use.
Skill bundles and scan evidence
Explore LOLSkills (opens in a new tab)Prevention testing
Turn living-off-the-land research into repeatable, ATT&CK-mapped tests to check whether your security controls block the techniques you care about.
Adversary playbooks and control validation
Explore Magic-Atomics (opens in a new tab)Detection engineering
Convert Sigma rules for your detection platform through a web interface powered by pySigma backends.
Sigma rule conversion
Explore Sigconverter (opens in a new tab)Bootloader intelligence
Investigate known malicious bootloaders and the research behind persistent threats that operate below the operating system.
Malicious bootloaders and research
Explore Bootloaders (opens in a new tab)Bring a finding, improve a detection, or help test a fix. Contributions help the next defender make a better decision.
View all projectsChoose the sources that fit your environment, then decide what runs, one policy at a time.
Free forever. No credit card required.