On September 23, 2026, an attacker spent roughly five hours poisoning two packages belonging to MemTensor, the company behind the MemOS operating system for AI memory. By the time a researcher flagged the issue on GitHub at 4:17 AM UTC, malicious versions were already sitting at the top of the npm and PyPI registries, ready to install for any developer who ran a plain `npm install` or `pip install MemoryOS` that morning.
The packages hit were `@memtensor/memos-cloud-openclaw-plugin` on npm and `MemoryOS` on PyPI. Security firm SafeDep, which flagged the incident through its threat intelligence monitoring, found that three npm versions, `0.1.21`, `0.1.23`, and `0.1.25`, and one PyPI version, `2.0.34`, all contained the same Go binary: a credential-harvesting implant the attacker internally called `sckit`, built under the module path `supplychain.local/campaign`.
How the Attacker Got Inside the Pipeline
The attacker did not need a zero-day. Instead, they exploited a well-understood weakness in how GitHub Actions jobs share environment state.
The OpenClaw plugin publishes to npm through a GitHub Actions release workflow that reads its publish token from a repository secret. The attacker, operating through a GitHub account called `Memtensor-AI`, pushed a short-lived branch named `sc/release-0.1.21-20260922-cloud`, made a three-line change to a validation script that runs earlier in the same job, then deleted the branch. They repeated this process five times between 00:48 and 02:03 UTC.
The change was precise: it wrote a `BASH_ENV` entry into `$GITHUB_ENV`, which is GitHub's mechanism for passing environment variables between steps. Because Bash reads the file named in `BASH_ENV` before running any non-interactive script, this let the attacker's shell script execute silently before the real publish step. That script called `collectStageZero()` from within the package itself, passed the `NPM_TOKEN` to the `sckit` binary, then deleted itself and exited with a failure code. The publish step failed visibly, so nothing appeared on npm from that run. The token was already gone.
The PyPI compromise used the same `BASH_ENV` trick but through a different entry point. The attacker pushed an unsigned commit to the MemOS repository that replaced the standard build backend in `pyproject.toml` with a custom wrapper called `sckit_poetry_build`. On import, that wrapper injected its own bridge script into the CI environment. The bridge ran only inside the PyPI upload action's container, captured `INPUT_PASSWORD` (the PyPI token), sent it to a server at `10729e014d0e.skyleen[.]fr`, and then exited cleanly. Two hours later, a follow-up commit removed the capture code, and the next tag push uploaded the fully malicious wheel to PyPI using MemTensor's own legitimate credentials.
What the Package Does After Install
The implant activates at runtime, not at install time, so `--ignore-scripts` offers no protection. In the npm plugin, it fires when the OpenClaw gateway starts and again on every memory recall. In the Python library, it starts the first time `configure_logging()` runs, which happens on nearly every import path. The binary launches detached in the background with no output.
Once running, `sckit` scans the entire home directory for credentials. Its target list, visible in its strings and symbol names, covers `.npmrc`, `.pypirc`, `.git-credentials`, `.netrc`, SSH private keys, HashiCorp Vault tokens, and Microsoft MSAL token caches. Two compiled regular expressions recognize both secret-like variable names and token format patterns for AWS, GitHub, npm, PyPI, HuggingFace, Slack, and Stripe. Collected data goes to subdomains of `skyleen[.]fr`, the campaign's control infrastructure, over encrypted channels using X25519 key exchange and XChaCha20-Poly1305.
The binary also carries worm logic. Functions named `findRepositories`, `prepareRemoteNode`, `prepareRemotePython`, and `recursivePublish` describe how it uses stolen credentials to inject itself into other repositories. It plants a GitHub Actions workflow named `runtime-update.yml` and a `.sckit/` directory into reachable projects, turning each victim into a potential carrier. The campaign configuration encodes an expiry date of late October 2026, suggesting the attacker planned a defined window of operation.
Developers Need to Act Now
Anyone who ran an affected version should treat every credential in their home directory as stolen. That includes cloud CLI tokens, SSH keys, and any `.env` files. SafeDep recommends pinning to `0.1.20` for the npm plugin and `2.0.33` for `MemoryOS`, killing any running `sckit` process, deleting the state directories at `$HOME/.openclaw/.cache/runtime` and `$HOME/.memos/.cache/runtime`, and checking any repository with push access for the `runtime-update.yml` workflow file.
The attack sits inside a larger pattern. The first half of 2026 alone produced 37 supply chain attack campaigns and 497 indexed malicious packages, which is 4.5 times the package volume of the entire preceding year. What separates this incident is the operational sophistication: the attacker used the target project's own CI pipeline as the delivery mechanism, left no workflow run logs behind, and built self-propagation directly into the implant. For maintainers who publish from CI, PyPI's trusted publishing removes long-lived tokens from the job entirely. Required reviewers on release environments would have blocked the MemTensor runs before they started.
The extortion group ShinyHunters hacked the dark web leak site run by Clop, one of the most active ransomware operations in the world, defaced it with their own branding, and is now threatening to put Clop through the same extortion process Clop runs on its corporate victims.
The attack happened Friday night, September 19. ShinyHunters found an unauthenticated file upload flaw in Grav CMS, the content management system Clop was running its leak site on, and used it to push a text file directly onto the server. The file read: "THIS SITE HAS BEEN PWN3D BY SHINYHUNTERES #Skids10p - Maybe don't try to threaten us next time." It also linked back to ShinyHunters' own Tor site. The file was confirmed live and downloadable directly from Clop's server.
Hours later, ShinyHunters said they had gone further. A visit to Clop's site showed the entire page replaced with ASCII art of Umbreon, the Pokemon ShinyHunters uses as its logo, and the line "rooting your systems since '19 ;)". The same Umbreon artwork had appeared when ShinyHunters defaced HackForums back in August 2020. Clop's defaced page was still live at the time of writing.
ShinyHunters claimed full access to the server and said they took source code, Grav CMS plugins, and everything stored in the server's /var/log directory, which typically holds authentication logs, system activity records, and the IP addresses of everyone who connected to it. They also claim to have pulled the private keys for Clop's Tor onion service. Those keys are what tie a .onion address to its server. With them, ShinyHunters could host a copy of Clop's site at the exact same onion URL, on infrastructure they control. "We have their onion keys. So if they kick us out it wouldn't matter at all because we control the private keys to host the same exact onion URL," the group said.
The plan is to post an extortion message on their own site and give Clop 72 hours to respond.
The defacement and the uploaded file are independently confirmed. The claims about stolen source code, server logs, and Tor private keys come only from ShinyHunters and have not been independently verified. Clop has not commented.
The dispute behind this attack goes back about a year. In August 2025, Clop quietly began exploiting a zero-day vulnerability in Oracle E-Business Suite, tracked as CVE-2025-61882, a server-side request forgery flaw that gave attackers remote access to enterprise systems without authentication. Oracle did not patch it until October 2025, after Mandiant confirmed active exploitation. By then, Clop had already sent mass extortion emails to executives at dozens of companies, including Cox Enterprises, The Washington Post, Logitech, Michelin, and Estee Lauder.
ShinyHunters says that exploit was originally theirs and that Clop used it without authorization. In October 2025, ShinyHunters, operating under the name "Scattered Lapsus$ Hunters," leaked the proof-of-concept publicly. Oracle confirmed it matched the exploit used in the Clop attacks. ShinyHunters said the leak was deliberate, intended to disrupt Clop's campaign and expose what had been taken from them.
What followed, according to ShinyHunters, was a direct threat from a Clop representative. "During the Oracle EBS campaign they ran and stole from me last year, someone from cl0p personally messaged me and said, and I quote (translated from Russian): I have more money than you and all of your people combined, I'll kill you soon," the group said. Those allegations have not been independently verified.
This is not the first time criminal groups have turned on each other. In March 2025, DragonForce defaced the leak sites of rival operations BlackLock and Mamona. Later in 2026, two groups called 0APT and KryBit hacked and leaked each other's operational data until both were left severely damaged. The difference in the Clop case is the scale of the target. Clop's leak site is the operational center of its entire extortion model, the platform it uses to name victims and apply public pressure when ransoms go unpaid. Losing control of it, and potentially the keys that anchor its onion address, is not a minor disruption.
ShinyHunters' own Tor site went offline shortly after the attack. No connection to Clop has been established.
The vulnerability, known as CVE-2026-20079, is given a maximum CVSS score 10.0, which makes it one of the most dangerous flaws impacting Cisco’s firewall management products.
The flaw was first disclosed in March 2026 by Cisco, but on September 9, Cisco updated its security advisory to confirm about the active exploitation in August that its Product Security Incident Response Team (PSIRT) became aware about. Cisco has advised users to update impacted systems immediately.
The flaw impacts the web interface of Cisco Secure Firewall Management Center Software. When an improper system process is created after the starting of the impacted device, it results in the flaw.
The threat actor does not require authentic credentials to abuse the vulnerability. A remote attacker can escape verification by sending specially tailored HTTP requests to a compromised FMC device.
Cisco has listed the problem as authentication bypass using a different channel or path, or CWE-288. As the flaw can be abused remotely without user interaction or verification, Cisco has given it a CVSS score of 10.0.
Vulnerable products
According to Cisco, regardless of device configuration, the flaw impacts Cisco Secure FMC Software and Cisco Security Cloud Control (SCC) Firewall Management.
Not vulnerable products
The following products are not impacted by the vulnerability:
The flaw could have severe impact for enterprises using Cisco Secure Firewall Management Center for managing their security infrastructure
If a threat actor gains root access, they may modify system configurations, install additional malware, use the infected management system as a base for future attacks and run malicious commands.
The vulnerability could have serious consequences for organizations using Cisco Secure Firewall Management Center to manage their security infrastructure.
An attacker who gains root access could potentially alter system configurations, execute malicious commands, install additional malware or use the compromised management system as a foothold for further attacks.
“To determine if this vulnerability may have been exploited, use the zgrep "package_info.*license" messages* CLI command in expert mode,” Cisco said.
According to Cisco, if organizations suspect exploit, they should reach out to the Cisco Technical Assistance Center (TAC) for help with recovery options.
Dutch e-commerce security firm Sansec found the vulnerability, called “StyleSmuggler,” which lets threat actors run malicious code on compromised servers without needing verification. According to Sansec, it found active attacks starting on September 4,2026 and posted the information after discovering that online stores were already being targeted.
StyeSmuggler is a severe remote code execution (RCE) vulnerability impacting present Magento variants. According to Sansec, the flaw successfully reproduced the attack on Magento Open Source variants 2.4.7, 2.4.8 and 2.4.9. This includes the latest 2.4.9 release.
For people who are not using Sansec Shield, they can temporarily stop Graphql until Adobe has issued an official patch. “The next Adobe security release is scheduled for Sept 8th, but it is yet unknown if that covers this bug,” Sansec said.
The flaw is notedly troubling as threat actors don’t require any password, admin account, or a username to exploit it. A successful compromise could give attackers the power to run code directly on the server hosting an online store.
Malicious code is injected into Magento's template system by StyleSmuggler. It can get around present safety precautions by utilizing the styles features. It operates in two phases:
Threat actors exploit GraphQL and Magento’s template functionality to install malicious code, The attack can then prompt the installed code via Magento’s internal processes, in the end letting them download and run malware on the infected server.
The attack is not limited to stores running old software, which is a troubling finding.
“The first victim ran 2.4.6-p15 with the July and August 2026 patches applied and security:patch-status clean,” Sansec said. In other words, even having the latest available security updates for the released could not stop the exploit of the new zero-day.
Another group, Disrex found the exploit, a Magento hosting and development company. “One of the Magento servers managed by Disrex was compromised only 50 minutes after the first confirmed StyleSmuggler attack worldwide,” it said.
No official patch has been released at the time of writing this article.
A Russian national has been extradited to the United States to face federal charges over an alleged malware campaign that targeted approximately 80,000 users of a freelance employment platform.
Searzhudin Tamirlanovich Aktulaev, 40, is accused of using hundreds of fraudulent accounts to distribute malicious Microsoft Excel attachments between June 2016 and November 2017. Prosecutors allege that the attachments downloaded remote-access malware capable of controlling victims’ computers and stealing information.
The indictment was filed under seal on June 1, 2021. Aktulaev was arrested in Cyprus in May 2025 and extradited to the United States on August 28, 2026, according to the US Department of Justice.
He appeared in federal court in San Francisco on August 31 and was remanded to federal custody. The indictment was unsealed the same day.
According to prosecutors, Aktulaev and his alleged co-conspirators exploited the messaging system of a well-known freelance employment technology company located in California’s Northern District.
The DOJ did not publicly identify the company.
The conspirators allegedly created approximately 255 fake user accounts and used them to send messages containing malicious Excel attachments to around 80,000 freelancers.
Opening an attachment prompted the recipient to enable or execute an embedded macro. If the user complied, the macro downloaded malware from the internet.
This distinction is important: the indictment alleges that malware was distributed to approximately 80,000 users, but the DOJ announcement does not establish that every recipient opened the attachment or became infected.
Freelancers can be particularly exposed to attachment-based attacks because their work routinely involves receiving documents from unfamiliar prospective clients. A spreadsheet presented as a project brief, financial record or work assignment may therefore appear consistent with an ordinary business request.
The indictment identifies two malware families allegedly used in the campaign: TVRAT and DarkVNC.
TVRAT, also known as TVSPY or TeamSpy, incorporated or abused components associated with TeamViewer, a legitimate remote-administration product. DarkVNC provided similar hidden remote-control capabilities using Virtual Network Computing technology.
Prosecutors allege that the malware allowed the conspirators to control infected computers and transfer stolen data to command-and-control servers.
The use of recognizable remote-access components can help criminals disguise malicious activity as legitimate administrative traffic. It can also make detection more difficult when organizations permit remote-support products in their environments.
The allegations do not indicate that TeamViewer or VNC Viewer participated in the operation. The case concerns malware that allegedly misused remote-administration technology.
Court allegations state that domains supporting the command-and-control infrastructure were purchased using virtual currency. At least one command-and-control domain was hosted in the United States, and thousands of computers infected with TVRAT reportedly connected back to it.
Investigators discovered a database on the command-and-control infrastructure containing information associated with thousands of victims.
The DOJ also said a shared document stored in an email account used during the alleged criminal activity contained e-commerce login credentials and personally identifiable information belonging to hundreds of people.
Prosecutors allege that information stolen through TVRAT and DarkVNC was collected from the command-and-control servers and used by Aktulaev and his co-conspirators to conduct fraud and other criminal activity.
Approximately half of the identified victims were located in the United States. The DOJ said many—not all—of those US victims were in the Northern District of California, where the federal case is being prosecuted.
The indictment charges Aktulaev with conspiracy, aggravated identity theft and transmitting code or commands that caused damage to protected computers. It also includes allegations involving wire fraud, unauthorized computer access and obtaining information or value from compromised systems.
The most serious listed offense, conspiracy to commit wire fraud, carries a maximum potential sentence of 20 years in prison. A charge involving intentional damage to protected computers carries a maximum of 10 years, while aggravated identity theft can result in a mandatory consecutive two-year sentence for each conviction.
These are maximum statutory penalties rather than a predicted sentence. Any punishment would depend on which charges, if any, result in conviction and the federal court’s consideration of applicable sentencing rules.
Aktulaev is scheduled to appear before US District Judge Donato on October 5, 2026, for a status conference.
The FBI investigated the case, which is being prosecuted by the National Security, Cyber, and Special Prosecutions Section of the US Attorney’s Office for the Northern District of California. The Justice Department’s Office of International Affairs secured the extradition from Cyprus.
The alleged campaign operated from 2016 to 2017. The indictment was filed in 2021, approximately four years after the campaign ended—not four years after Aktulaev’s arrest.
Aktulaev was arrested in Cyprus in May 2025 and extradited more than a year later. The DOJ has not explained why the indictment remained under seal or provided details about the extradition proceedings.
Because the case is at the indictment stage, all described conduct remains an allegation. Aktulaev is presumed innocent unless prosecutors prove the charges beyond a reasonable doubt.
An international public-private operation has disrupted Sality, a peer-to-peer botnet that remained active for more than two decades, by seizing domains and redirecting infected computers away from infrastructure controlled by its operator.
The coordinated action took place on August 31, 2026, and involved authorities in the United States, Bulgaria, Hungary and Romania. CrowdStrike and the Shadowserver Foundation provided technical assistance, while Europol and Eurojust supported the international coordination.
According to the US Department of Justice, the DOJ, FBI and the Defense Criminal Investigative Service seized Sality-linked domains in the United States. European authorities took action against additional domains hosted in Bulgaria, Hungary and Romania.
CrowdStrike’s Counter Adversary Operations team simultaneously carried out a peer-to-peer sinkholing operation designed to separate infected computers from the botnet’s operator.
The action disabled Sality’s current command channel, but it did not automatically remove malware from compromised computers.
First observed in 2003, Sality began as a polymorphic file-infecting malware family. It attached malicious code to executable files and could spread through network shares, removable drives and file-sharing systems.
Sality eventually developed into a decentralized botnet in which infected computers communicated directly with one another. Unlike a conventional botnet with one central command-and-control server, Sality’s peer-to-peer architecture did not present investigators with a single server that could be seized to disable the entire operation.
CrowdStrike said two incompatible Sality networks, known as versions 3 and 4, remained active until the disruption. Although they used the same underlying codebase and were operated by the same threat actor, the networks used different protocol versions and cryptographic keys.
The botnet’s main function was to deliver additional malicious software. During its long history, Sality distributed malware associated with credential theft, spam, proxy services, network exploitation and distributed denial-of-service attacks.
For approximately the past eight years, CrowdStrike said its primary payload was EggJagger, a clipboard-hijacking tool that monitored devices for copied cryptocurrency wallet addresses. When it detected a Bitcoin or Ethereum address, the malware replaced it with an address controlled by the attacker.
CrowdStrike estimates that at least $150,000 in cryptocurrency was stolen through EggJagger, although other malware distributed through Sality may have produced additional criminal revenue.
CrowdStrike’s technical account of the disruption states that Sality enabled its operator to distribute malicious payloads to more than 33,000 infected computers worldwide at the time of the operation.
Historical figures are considerably larger. Europol reported that Sality provided access to as many as one million infected machines at its peak. More than 11 million unique IP addresses have been connected to its infrastructure over its lifetime.
The 11 million figure should not be interpreted as the number of simultaneously infected devices. Individual machines can use different IP addresses over time, and the total covers years of recorded activity.
Europol has supported Sality-related investigations since 2017. The agency said international partners held weekly operational calls in the weeks before the latest action to coordinate infrastructure seizures and the technical disruption.
The operation targeted the peer lists that Sality-infected computers used to locate other machines in the botnet.
Each infected computer maintained a limited list of publicly reachable “super peers,” which formed the backbone of the P2P network. Approximately every 40 minutes, the malware checked whether those peers remained available. Responsive peers gained reputation, while unresponsive entries were gradually removed.
CrowdStrike found that the Sality protocol did not authenticate computers joining the network. Any publicly reachable system that completed the required handshake could be accepted as a legitimate peer.
Defenders used this weakness to manipulate the botnet’s peer lists. Legitimate Sality peers were invalidated and replaced with sinkhole nodes operated by CrowdStrike. As the process continued, infected computers lost contact with the criminal network and began communicating with defender-controlled infrastructure instead.
Computers located behind firewalls or network address translation could not always be contacted directly. However, when those devices initiated their routine communications with a sinkhole, their peer lists could also be purged, isolating them from the operator.
Authorities and industry partners also acted against websites hosting Sality’s payloads. Taking those locations offline prevented infected computers with older download instructions from retrieving additional malware during the transition.
CrowdStrike tracks the criminal actor associated with Sality as SALTY SPIDER. The company assesses that the group likely operates from Russia’s Republic of Bashkortostan, near the border with Kazakhstan.
This remains a company attribution rather than a publicly established legal finding. The DOJ and Europol announcements did not identify an individual operator, announce an arrest or disclose criminal charges connected to the disruption.
Although the operator has lost the ability to issue new instructions through the disrupted network, Sality remains installed on affected computers. Additional malware previously delivered by the botnet may also remain active.
CrowdStrike said Sality-infected systems now communicate with its sinkhole infrastructure. The company published a defanged lighthouse IP address, the botnet’s final payload URLs and YARA rules that security teams can use to identify active infections.
Shadowserver is working with internet service providers and national Computer Security Incident Response Teams to identify affected organizations, notify victims and support remediation.
Organizations that detect a Sality infection should isolate the affected device, inspect it for secondary malware and remove or rebuild compromised systems. Security teams should also examine network shares and removable media that may contain infected executable files.
Passwords and other credentials used on an infected device should be considered exposed. However, credential changes should be performed from a clean system after the malware has been removed.
The operation has disabled Sality’s existing command channel and prevented it from distributing new payloads through the disrupted infrastructure. Its long-term impact will now depend on whether remaining infections are found and remediated before the operator can attempt to rebuild part of the network.