Threat research

Blogs Cybersecurity News

Scanners Turn Attack Vector as TrivyScanner Hijacked via GitHub Actions Tags

Attackers Targeted SSH keys, Cloud Tokens & API secrets in CI/CD Pipelines; Highlights Securing CI/CD Pipelines

In latest vulnerability discovery Aqua Security revealed HackerBot-claw bot hijacked 75 of 76 GitHub Actions tags for its Trivy vulnerability scanner. The HackerBot-claw first distributed credential-stealing malware through the widely used security tool for the second time in a one month.

Malicious code rode alongside legitimate scans, targeting SSH keys, cloud tokens and API secrets in CI/CD pipelines. Security researcher Paul McCarty was the first to warn publicly that Trivy version 0.69.4 had been backdoored, with malicious container images and GitHub releases published to users.

Attack module on Trivy

When it comes to workflow it has been observed that more then 10,000 GitHub workflow files rely on trivy-action. Attackers can leverage this pipeline and pull versions during the attack window which are affected and carry sensitive credentials exfiltrated.

Attackers compromised the GitHub Action by modifying its code and retroactively updating version tags to reference a malicious commit. This permitted data used in CI/CD workflows to be printed in GitHub Actions build logs and finally leaking credentials.

A self-propagating npm worm compromised 47 packages, extending the blast radius into the broader JavaScript ecosystem.

Aqua Security disclosed in a GitHub Discussion that the incident stemmed from incomplete containment of an earlier March 1 breach involving a hackerbot-claw bot.

  • Attackers swapped the entrypoint.sh in Trivy’s GitHub Actions with a 204-line script that prepended credential-stealing code before the legitimate scanner.
  • Lines 4 through 105 contained the infostealer payload, while lines 106 through 204 ran Trivy as normal.
  • This made difficult  to detect during routine scans.

TeamPCP preserved normal scan functionality to avoid triggering CI/CD failures as detection now will require cryptographic verification of commit signatures .

For defenders, traditional CI/CD monitoring, which watches for build failures or unexpected output, can no longer catch supply-chain compromises that deliberately maintain normal behavior.

Organizations relying on Trivy or similar open-source security tools are facing attacks from the very scanners meant to protect their pipelines can become the attack vector. Only cryptographic provenance checks can distinguish legitimate releases from poisoned ones.

As per security researchers once inside a pipeline, the malicious script scanned memory regions of the GitHub Actions Runner.

Github Compromise

The attack appears to have been accomplished via the compromise of the cx-plugins-releases (GitHub ID 225848595) service account, as that is the identity involved in publishing the malicious tags. 

Credentials exfiltrated during the initial incident were used last week in a new supply chain attack that targeted not only the Trivy package but also trivy-action and setup-trivy, Trivy’s maintainers have confirmed in a March 21 advisory.

Key Findings b Wiz Research

  • According to Wiz, the attack appears to have been carried out via the compromise of the “cx-plugins-releases” service account, with the attackers with malicious container images and GitHub releases published to users.
  • The second stage extension is activated and the malicious payload checks whether the victim has credentials from cloud service providers such as GitHub, AWS, Google Cloud, and Microsoft Azure.
  • When credentials if they are detected, it proceeds to fetch a next-stage payload from the same domain (“checkmarx[.]zone”).

“The payload attempts execution via npx, bunx, pnpx, or yarn dlx. This covers major JavaScript package managers,” Wiz researchers Rami McCarthy, James Haughom, and Benjamin Read said. “The retrieved package contains a comprehensive credential stealer.

Harvested credentials are then encrypted, using the keys as elsewhere in this campaign, and exfiltrated to ‘checkmarx[.]zone/vsx’ as tpcp.tar.gz.”

Conclusion: Aqua Security urged affected users to “treat all pipeline secrets as compromised and rotate immediately.” 

Organizations that ran any version of trivy-action, setup-trivy, or Trivy v0.69.4 during the attack window should audit their CI/CD logs for unexpected network connections to scan.aquasecurtiy[.]org and check whether any tpcp-docs repositories were created under their GitHub accounts.

With three major tag-hijacking incidents in 12 months, Wiz security researcher Rami McCarthy recommended that organizations “pin GitHub Actions to full SHA hashes, not version tags.”

Sources: Trivy Breached Twice in a Month via GitHub Actions

Blogs Cybersecurity News

AI Lazed Attacks Targets 600+FortiGate Firewalls; Amazon; What Makes AI a Powerful tool for Hackers

AI Lazed Threat Actors Compromised 600 +Fortinet FortiGate firewalls in 55 countries

Continue Reading
Security Advisory

Multiple Critical Vulnerabilities in Citrix NetScaler ADC/Gateway 

Security Advisory: Multiple vulnerabilities have been discovered in NetScaler ADC (formerly Citrix ADC) and NetScaler Gateway One Actively Exploited in Wild .

Citrix credited Jimi Sebree of Horizon3.ai, Jonathan Hetzer of Schramm & Partnerfor and Francois Hammerli for discovering and reporting the vulnerabilities.

Severity Critical 
CVSS Score 9.2 
CVEs CVE-2025-7775, CVE-2025-7776, CVE-2025-8424 
POC Available No 
Actively Exploited Yes 
Exploited in Wild Yes 
Advisory Version 1.0 

Overview 
A critical zero-day vulnerability, tracked as CVE-2025-7775, puts over 28,200 Citrix instances at risk worldwide.

This flaw allows attackers to run malicious code on affected systems without authentication. The issue is actively being exploited in the wild and immediate action is needed to secure systems.  Another two flaws were fixed in the latest updates.  

Vulnerability Name CVE ID Product Affected Severity 
Memory overflow vulnerability leading to RCE CVE-2025-7775 NetScaler ADC & Gateway 9.2 
Memory overflow vulnerability leading to unpredictable behavior CVE-2025-7776 NetScaler ADC & Gateway 8.8 
Improper access control on the NetScaler Management Interface CVE-2025-8424 NetScaler ADC & Gateway 8.7 

Technical Summary 

The NetScaler ADC and NetScaler Gateway appliances are affected by multiple critical vulnerabilities that pose significant risks ranging from Remote Code Execution (RCE) and Denial of Service (DoS) to improper access control.

These include memory overflow flaws in configurations such as VPN virtual servers, load balancing virtual servers using IPv6 or DBS IPv6 services, and misconfigurations involving PCoIP profiles. Additionally, the management interface is exposed due to weak access control mechanisms, which could allow unauthorized administrative access if attackers reach key management IP addresses like NSIP or SNIP. CISA has added one vulnerability (CVE-2025-7775) to its Known Exploited Vulnerabilities (KEV) Catalog and strongly urges organizations to apply patches immediately to prevent active exploitation. 

CVE ID System Affected  Vulnerability Details Impact 
CVE-2025-7775  NetScaler ADC & Gateway  A critical memory overflow vulnerability in NetScaler ADC and Gateway that can lead to Remote Code Execution or DoS when configured as a Gateway (e.g., VPN, ICA Proxy, CVPN, RDP Proxy), AAA virtual server, or LB virtual server using IPv6 or DBS IPv6 services including CR virtual servers of type HDX. Remote Code Execution or DoS  
CVE-2025-7776  NetScaler ADC & Gateway A memory overflow vulnerability under analysis, currently known to cause unpredictable system behavior and potential DoS when a PCoIP Profile is bound to a Gateway-configured NetScaler instance (VPN, ICA Proxy, CVPN, RDP Proxy), Erroneous behavior and DoS 
CVE-2025-8424 NetScaler ADC & Gateway An improper access control vulnerability on the NetScaler Management Interface, allowing unauthorized access when attackers can reach management IPs (NSIP, Cluster Management IP, local GSLB Site IP, or SNIP with Management Access), affecting NetScaler ADC and Gateway appliances. Unauthorized access 

Recommendations 

NetScaler ADC and NetScaler Gateway to install the relevant updated versions as soon as possible.  

  • NetScaler ADC and NetScaler Gateway 14.1-47.48 and later releases 
  • NetScaler ADC and NetScaler Gateway 13.1-59.22 and later releases of 13.1 
  • NetScaler ADC 13.1-FIPS and 13.1-NDcPP 13.1-37.241 and later releases of 13.1-FIPS and 13.1-NDcPP 
  • NetScaler ADC 12.1-FIPS and 12.1-NDcPP 12.1-55.330 and later releases of 12.1-FIPS and 12.1-NDcPP 

Here are some other recommendations below 

  • Monitor systems for unusual activity or unauthorized changes. 
  • Limit access to Citrix instances from untrusted networks. 
  • Use firewalls to block suspicious traffic targeting Citrix instances. 

Conclusion: 

Combined with additional high-severity vulnerabilities the overall threat landscape demands immediate attention. Organizations are strongly urged to apply the latest patches, restrict access to management interfaces and closely monitor for signs of compromise. Delayed action could result in significant operational and security impacts. 

The active exploitation of CVE-2025-7775 highlights a critical security threat affecting multiple NetScaler ADC and Gateway instances globally. This zero-day confirmed exploitation in the wild poses a severe risk of Remote Code Execution and service disruption.

References: 

  

Security Advisory

Microsoft IIS Web Deploy RCE Vulnerability Allows Authenticated Remote Code Execution 

Summary of Vulnerability in Microsoft Web Deploy 4.0 (CVE-2025-53772) revels critical security flaw that could be exploited by authenticated attackers to execute code on affected systems. This is the bug disclosed on August 12, 2025, with a CVSS score of 8.8, indicating high severity.

Severity High 
CVSS Score 8.8 
CVEs CVE-2025-53772 
POC Available No 
Actively Exploited No 
Exploited in Wild No 
Advisory Version 1.0 

Overview 

A vulnerability in Microsoft Web Deploy 4.0 (CVE-2025-53772) allows authenticated attackers to remotely execute arbitrary code on affected systems.

The issue arises from the insecure deserialization of untrusted data. Due to its low privilege requirements and lack of user interaction, this flaw poses a significant threat, especially in enterprise deployment environments. 

Vulnerability Name CVE ID Product Affected Severity Fixed Version 
​ Web Deploy Remote Code Execution via Deserialization  CVE-2025-53772 Microsoft Web Deploy 4.0  High  10.0.2001 or later 

Technical Summary 

The vulnerability stems from insecure deserialization of untrusted data (CWE-502), allowing remote attackers to craft malicious HTTP requests that trigger code execution on the web server. This flaw enables remote code execution (RCE) under specific conditions, where the attacker must have authenticated access and network connectivity.

The attack is network-based, requires only low-privilege access and does not rely on user interaction. Successful exploitation can result in a high impact on confidentiality, integrity and availability of the affected system. As of the time of publication, no public exploit has been reported and the exploit maturity is considered unproven. 

CVE ID CVSS Score System Affected  Vulnerability Details Impact 
CVE-2025-53772 8.8 Microsoft Web Deploy 4.0 Web Deploy deserializes untrusted input, allowing remote attackers to execute arbitrary code. Remote Code Execution 

Recommendations: 

Here are some recommendations below 

  • Apply Microsoft Web Deploy version 10.0.2001 or latest version. 
  • Limit access to Web Deploy endpoints to trusted IP ranges or internal networks only. 
  • Audit logs for unusual HTTP POST activity to Web Deploy endpoints. 

Conclusion: 
While CVE-2025-53772 has not yet been publicly exploited, the nature of the flaw and the ease of attack (low privileges, no user interaction) significantly increases the risk of widespread exploitation, particularly in enterprise deployment environments.

Organizations using Microsoft Web Deploy 4.0 should update and apply the latest patch without delay.

This vulnerability affects Web Deploy 4.0 and requires low privileges to exploit, making it particularly concerning for organizations that use this deployment tool in their infrastructure. The vulnerability allows an authenticated attacker to exploit the system via low-complexity network-based attacks. 

References: 

Blogs Security Advisory

New Cyberattack Methodology ‘Man in Prompt’, User’s at Risk, Target-AI Tools

AI tools like ChatGPT, Google Gemini and others being afflicted by malicious actors via injecting harmful instructions into leading GenAI tools. These were overlooked previously and attack methodology targets the browser extensions installed by various organizations.

The attack methodology named as ‘Man in Prompt’, exercise its attack with new class exploit targeting the AI tools as per LayerX’s researchers.

As per the research any browser extension, even without any special permissions, can access the prompts of both commercial and internal LLMs and inject them with prompts to steal data, exfiltrate it and cover their tracks. 

The exploit has been tested on all top commercial LLMs, with proof-of-concept demos provided for ChatGPT and Google Gemini. 

The question is how do they impact Users & organizations at large & how does the AI tools function within web browsers?

For organizations the implications can be high then expected as AI tools are most sought after and slowly organization across verticals are relying on AI tools.

The LLMs used and tested on many organizations are mostly trained ones. They carry huge data set of information which are mostly confidential and possibility of being vulnerable to such attack rises .

The attack methodology named as ‘Man in Prompt’, exercise its attack with new class exploit targeting the AI tools as per LayerX’s researchers. As per the research any browser extension, even without any special permissions, can access the prompts of both commercial and internal LLMs and inject them with prompts to steal data, exfiltrate it, and cover their tracks. 

The attack methodology named as ‘Man in Prompt’, exercise its attack with new class exploit targeting the AI tools as per LayerX’s researchers. As per the research any browser extension, even without any special permissions, can access the prompts of both commercial and internal LLMs and inject them with prompts to steal data, exfiltrate it, and cover their tracks. 

LayerX researcher termed this type of attack as ‘hacking copilots’ that are equipped to steal organizational information.

The prompts given are a part of the web page structure where input fields are known as the Document Object Model, or DOM. So virtually any browser extension with basic scripting access to the DOM can read or alter what users type into AI prompts, even without requiring special permissions.

Bad actors can use compromised extensions to carry out activities including manipulating a user’s input to the AI.

  • Perform prompt injection attacks, altering the user’s input or inserting hidden instructions.
  • Extract data directly from the prompt, response, or session.
  • Compromise model integrity, tricking the LLM into revealing sensitive information or performing unintended actions

Understanding the attack scenario

Proof-of-concept attacks against major platforms

For ChatGPT, an extension with minimal declared permissions could inject a prompt, extract the AI’s response and remove chat history from the user’s view to reduce detection.

LayerX implemented an exploit that can steal internal data from corporate environments using Google Gemini via its integration into Google Workspace.

Over the last few months, Google has rolled out new integrations of its Gemini AI into Google Workspace. Currently, this feature is available to organizations using Workspace and paying users.

Gemini integration is implemented directly within the page as added code on top of the existing page. It modifies and directly writes to the web application’s Document Object Model (DOM), giving it control and access to all functionality within the application

These platforms are vulnerable to  any exploit which Layer X researchers showcased that without any special permissions shows how practically any user is vulnerable to such an attack. 

Threat mitigation

These kind of attacks creates a blind spot for traditional security tools like endpoint Data Loss Prevention (DLP) systems or Secure Web Gateways, as they lack visibility into these DOM-level interactions. Blocking AI tools by URL alone also won’t protect internal AI deployments.

LayerX advises organisations to adjust their security strategies towards inspecting in-browser behaviour.

Key recommendations include monitoring DOM interactions within AI tools to detect suspicious activity, blocking risky extensions based on their behavior rather than just their listed permissions, and actively preventing prompt tampering and data exfiltration in real-time at the browser layer.

(Source: https://layerxsecurity.com/blog/man-in-the-prompt-top-ai-tools-vulnerable-to-injection/)

Security Advisory

Windows Update Stack Privilege Escalation Vulnerability (CVE-2025-21204) – PoC Released  

The flaw, disclosed by researchers at Cyberdom Blog, poses a significant risk to millions of Windows users and organizations relying on windows.

OEM Windows 
Severity HIGH 
CVSS Score 7.8 
CVEs CVE-2025-21204 
POC Available Yes 
Actively Exploited No 
Exploited in Wild No 
Advisory Version 1.0 

Overview 

A high-severity vulnerability in the Windows Update Stack, CVE-2025-21204, enables local attackers to escalate privileges to SYSTEM level by exploiting trusted path abuse through symbolic links. The flaw affects various versions of Windows 10, Windows 11, and Windows Server.

A working proof-of-concept (PoC) exploit has been publicly released by security researcher Elli Shlomo, increasing the urgency to patch. The issue is addressed in the April 2025 cumulative update KB5055523. 

Vulnerability Name CVE ID Product Affected Severity CVSS Score 
​Windows Update Stack Privilege Escalation  CVE-2025-21204 Windows  HIGH  7.8 

Technical Summary 

The vulnerability lies in how Windows Update processes such as MoUsoCoreWorker.exe and UsoClient.exe, which run with SYSTEM privileges, handle directory junctions. Attackers can delete the legitimate Tasks directory under C:\ProgramData\Microsoft\UpdateStack and replace it with a symbolic link pointing to an attacker-controlled path. This allows the execution of arbitrary code as SYSTEM without triggering traditional security mechanisms.

A public PoC developed by Elli Shlomo demonstrates this exploit using only native Windows features—no external binaries or code injection required. 

This opens the door for a range of attacks, including installing persistent malware, disabling security tools, or accessing sensitive data.

CVE ID System Affected Vulnerability Details Exploit Prerequisites Impact 
  CVE-2025-21204  Windows 10 (10.0.10240.0 < 10.0.10240.20978, etc.), Windows 11, Server Misuse of NTFS junctions allows local attackers to redirect C:\ProgramData\Microsoft\UpdateStack\Tasks to attacker-controlled locations. SYSTEM-level update processes follow these junctions and execute unauthorized code. Attackers must have local access and limited user privileges; no user interaction required   Local privilege escalation, Code execution 

Source: Cyberdom 

Recommendations: 

  • Apply the April 2025 cumulative update (KB5055523) immediately. 
  • Restrict ACLs on C:\ProgramData\Microsoft\UpdateStack. 
  • Use AppLocker or WDAC to block symbolic link creation in sensitive directories. 
  • Monitor file operations involving UpdateStack and inetpub, regardless of IIS presence. 
  • Detect attempts to create NTFS junctions targeting update directories. 

Conclusion: 
CVE-2025-21204 is an example of a rather low-level and impactful threat doing trusted path abuse rather than complex memory corruption. This vulnerability demonstrates how attackers will exploit trust assumptions built into the operating system via native components.

The only defenses available are to immediately patch and harden directory access controls to stop this low-level and minimally visible localized privilege escalation. 

References: 


 

Security Advisory

Phishing Crusade Targeted approx 12,000 GitHub Repositories; Victims directed to “gitsecurityapp”

A large-scale phishing campaign has targeted nearly 12,000 GitHub repositories with phony security alerts, reported BleepingComputers.

The alerts, opened as issues on the repositories, inform users of unauthorized login attempts and provide links to change their passwords, review active sessions, or set up MFA.

If a user clicks any of these links, they’ll be taken to a GitHub authorization page for an OAuth app that will grant the attacker control of the account.

The campaign is ongoing, though GitHub appears to be responding to the attacks.

Users were directed to all links within the message to a GitHub authorization page for a malicious OAuth application called “gitsecurityapp.” If authorized, the app grants attackers full control over the user’s account and repositories, including the ability to delete repositories, modify workflows, and read or write organization data.

This consistent messaging across all affected repositories aims to create a sense of urgency and panic, prompting developers to take immediate action.

The fraudulent alert directs users to update their passwords, review active sessions, and enable two-factor authentication. However, these links lead to a GitHub authorization page for a malicious OAuth app named “gitsecurityapp.”

Upon authorization, an access token is generated and sent to various web pages hosted on onrender.com, granting the attacker full control.

(Image courtesy: Bleeping Computers)

The attack, which was first detected on March 16, remains active, though GitHub appears to be removing affected repositories.

Pointers Developers to take key inputs from this incident.

Last week, a supply chain attack on the tj-actions/changed-files GitHub Action caused malicious code to write CI/CD secrets to the workflow logs for 23,000 repositories.

If those logs had been public, then the attacker would have been able to steal the secrets.

The tj-actions developers cannot pinpoint exactly how the attackers compromised a GitHub personal access token (PAT) used by a bot to perform malicious code changes as per threat researchers.

Key pointers for User saftey:

  • For users who have mistakenly authorized the malicious OAuth app revoking access to suspicious OAuth apps through GitHub’s settings.
  • Affected users should review their repository workflows, check for unauthorized private gists, and rotate their credentials to prevent further damage.
  • This attack highlights the increasing threat of phishing campaigns targeting GitHub users.
  • As GitHub continues to investigate and respond, developers must remain vigilant and verify any security alerts before taking action.
  • Rotate your credentials and authorization tokens.

 Wiz suggests that potentially impacted projects run this GitHub query to check for references to reviewdog/action-setup@v1 in repositories.

If double-encoded base64 payloads are found in workflow logs, this should be taken as a confirmation their secrets were leaked.

Developers should immediately remove all references to affected actions across branches, delete workflow logs, and rotate any potentially exposed secrets.

(Sourece: Bleeping computers)

Scroll to top