---METADATA---
VENDOR: Smallstep
PRODUCT: Step CA
AFFECTED_VERSIONS: < 0.30.0
CONFIDENCE: high
MISSING: none
---END_METADATA---

Description Summary:
Smallstep Step CA versions below 0.30.0 are vulnerable to an improper authentication flaw in the SCEP UpdateReq function, allowing unauthenticated attackers to issue certificates.

Executive Summary:
A critical authentication bypass in Smallstep Step CA allows unauthenticated attackers to issue unauthorized digital certificates, potentially undermining the entire PKI infrastructure.

Vulnerability Details

CVE-ID: CVE-2026-30836

Affected Software: Smallstep Step CA

Affected Versions: < 0.30.0

Vulnerability: This vulnerability is caused by a failure to properly authenticate requests during the SCEP UpdateReq process. An unauthenticated attacker can exploit this to perform unauthorized certificate issuance, which violates the fundamental security assumptions of the certificate authority.

Business Impact

The ability for an unauthenticated actor to issue certificates poses a catastrophic risk to the organization's PKI infrastructure. With a CVSS score of 10.0, this vulnerability could allow attackers to perform man-in-the-middle attacks, impersonate legitimate services, or decrypt secure communications, leading to total loss of trust in the system.

Remediation Plan

Immediate Action: Update Smallstep Step CA to version 0.30.0 or later to ensure proper authentication is enforced for all certificate requests.

Proactive Monitoring: Audit existing certificate issuance logs for any unauthorized or unexpected certificate requests that occurred prior to the update.

Compensating Controls: Restrict access to the SCEP endpoint at the network level using firewalls or mutual TLS (mTLS) to ensure only authorized entities can interact with the service.

Exploitation Status

Public Exploit Available: False

Analyst Notes: As of Mar 19, 2026, there is no public information indicating active exploitation. However, the criticality of this vulnerability (CVSS 10) indicates that it should be mitigated immediately to prevent potential large-scale security breaches.

Analyst Recommendation

This vulnerability represents the highest level of risk to security infrastructure. Organizations must upgrade to version 0.30.0 immediately and conduct a thorough audit of their certificate store to ensure no unauthorized certificates have been issued.

---METADATA---
VENDOR: PinchTab
PRODUCT: PinchTab HTTP Server
AFFECTED_VERSIONS: See vendor advisory for affected versions
---END_METADATA---

Description Summary:
PinchTab, an HTTP server for AI agents, contains a vulnerability that could allow for unauthorized control over Chrome browser instances.

Executive Summary:
A security flaw in the PinchTab HTTP server could allow attackers to hijack AI-driven browser sessions, leading to unauthorized data access or malicious browser activity.

Vulnerability Details

CVE-ID: CVE-2026-30834

Affected Software: PinchTab HTTP Server

Affected Versions: See vendor advisory for affected versions

Vulnerability: PinchTab is a standalone HTTP server designed to give AI agents control over Chrome. The vulnerability likely involves an authentication bypass or insecure API endpoint that allows an attacker to intercept or issue commands to the browser instance controlled by the AI.

Business Impact

The CVSS score of 7.5 indicates a High severity. If exploited, an attacker could steal session cookies, access sensitive web accounts, or use the hijacked browser to perform actions on behalf of the user, resulting in a significant breach of privacy and potential financial loss.

Remediation Plan

Immediate Action: Update the PinchTab server software to the latest version and ensure that all AI agent communication is encrypted and authenticated.

Proactive Monitoring: Monitor the HTTP server logs for unexpected connections or commands that do not originate from authorized AI agent sources.

Compensating Controls: Restrict access to the PinchTab HTTP server to the local host or a dedicated, secure management network using firewall rules.

Exploitation Status

Public Exploit Available: false

Analyst Notes: As of March 9, 2026, there is no public information indicating active exploitation of this vulnerability. As AI integration grows, these types of bridge applications will become increasingly high-value targets.

Analyst Recommendation

The intersection of AI and browser control presents a unique security risk. It is imperative to apply the primary remediation immediately to prevent attackers from gaining control over automated browser sessions and the sensitive data they handle.

---METADATA---
VENDOR: Soft Serve
PRODUCT: Soft Serve Git Server
AFFECTED_VERSIONS: 0.6.0 to before 0.11.4
---END_METADATA---

Description Summary:
An authenticated SSRF vulnerability in Soft Serve allows SSH users to force the server to make requests to internal IP addresses, potentially exposing private services.

Executive Summary:
Authenticated users of Soft Serve can exploit a Server-Side Request Forgery (SSRF) flaw to gain unauthorized access to internal network services and sensitive metadata.

Vulnerability Details

CVE-ID: CVE-2026-30832

Affected Software: Soft Serve

Affected Versions: 0.6.0 to before 0.11.4

Vulnerability: An authenticated SSH user can trigger the repo import command with a crafted --lfs-endpoint URL. This allows the attacker to force the server to perform blind HTTP requests to internal/private IP addresses, which can be further exploited to read internal service data.

Business Impact

While the attack requires authentication, it allows users to bypass network segmentation and interact with internal APIs, cloud metadata services, or other private resources. This could lead to sensitive data exposure or further exploitation of internal systems. The CVSS score of 9.1 indicates a critical risk to internal infrastructure.

Remediation Plan

Immediate Action: Update Soft Serve to version 0.11.4 or later to implement proper validation of LFS endpoints.

Proactive Monitoring: Inspect server logs for repo import commands targeting internal IP ranges (e.g., 10.x.x.x, 192.168.x.x, 169.254.169.254).

Compensating Controls: Use network-level firewalls or egress filtering to prevent the Git server from initiating connections to sensitive internal subnets.

Exploitation Status

Public Exploit Available: true

Analyst Notes: As of Mar 7, 2026, this vulnerability is documented as a multi-stage SSRF. Although the initial request is blind, it can be chained into a full read vulnerability using a malicious LFS server.

Analyst Recommendation

Administrators should prioritize updating Soft Serve to version 0.11.4. Even though authentication is required, the ability for a user to probe and exploit the internal network represents a significant security failure that must be addressed promptly.

---METADATA---
VENDOR: GStreamer
PRODUCT: GStreamer
AFFECTED_VERSIONS: See vendor advisory for specific affected versions
---END_METADATA---

Description Summary:
GStreamer's rtpqdm2depay component is vulnerable to an Out-Of-Bounds Write, which could allow an unauthenticated attacker to execute remote code.

Executive Summary:
A critical Out-Of-Bounds Write vulnerability in GStreamer's RTP processing module enables unauthenticated remote code execution, threatening the total compromise of affected multimedia systems.

Vulnerability Details

CVE-ID: CVE-2026-3083

Affected Software: GStreamer GStreamer

Affected Versions: See vendor advisory for specific affected versions

Vulnerability: This vulnerability is an Out-Of-Bounds (OOB) Write located in the rtpqdm2depay element. An unauthenticated remote attacker can exploit this flaw by sending specially crafted RTP packets to a system using GStreamer to process media streams.

Business Impact

The potential for Remote Code Execution (RCE) carries a CVSS score of 8.8. A successful exploit allows an attacker to gain full control over the host system, leading to data theft, installation of malware, or use of the system as a pivot point for lateral movement within the corporate network.

Remediation Plan

Immediate Action: Apply the latest GStreamer security patches immediately, specifically targeting the plugins-good or relevant library packages containing the rtpqdm2depay element.

Proactive Monitoring: Use network intrusion detection systems (IDS) to monitor for malformed RTP traffic or unexpected crashes in media processing services.

Compensating Controls: Disable the rtpqdm2depay plugin if it is not required for business operations, or utilize sandboxing technologies to isolate GStreamer processes.

Exploitation Status

Public Exploit Available: false

Analyst Notes: As of March 18, 2026, there is no public information indicating active exploitation of this vulnerability. However, RCE vulnerabilities in widely used media frameworks are highly prized by threat actors and often lead to rapid exploit development.

Analyst Recommendation

This vulnerability represents a significant risk to any infrastructure performing real-time media processing. Immediate patching is required to mitigate the risk of unauthenticated remote code execution.

---METADATA---
VENDOR: FlowiseAI
PRODUCT: Flowise
AFFECTED_VERSIONS: See vendor advisory for affected versions
---END_METADATA---

Description Summary:
A high-severity vulnerability in the Flowise LLM flow builder could allow attackers to compromise customized large language model flows.

Executive Summary:
Flowise contains a high-severity security vulnerability that could lead to the unauthorized manipulation or compromise of LLM-based application workflows.

Vulnerability Details

CVE-ID: CVE-2026-30823

Affected Software: Flowise

Affected Versions: See vendor advisory for affected versions

Vulnerability: The vulnerability affects the Flowise drag-and-drop interface used for building LLM flows. Based on the CVSS score of 8.8, the flaw likely permits an attacker to intercept or modify the logic of AI applications.

Business Impact

Successful exploitation could lead to "prompt injection" at scale, data exfiltration from connected vector databases, or the hijacking of AI-driven business processes. The CVSS score of 8.8 highlights the significant risk to the integrity of AI infrastructure.

Remediation Plan

Immediate Action: Update Flowise to the latest version immediately to resolve the identified security flaw.

Proactive Monitoring: Audit LLM flow configurations for unauthorized changes and monitor API logs for unusual request patterns.

Compensating Controls: Ensure that Flowise is deployed within a secure network perimeter and that all users are authenticated via a centralized identity provider.

Exploitation Status

Public Exploit Available: false

Analyst Notes: As of March 8, 2026, no active exploitation has been reported. As LLM frameworks become more prevalent, they are increasingly targeted by researchers and threat actors alike.

Analyst Recommendation

Apply the vendor's security updates without delay. AI orchestration tools often have broad access to internal data sources, making their security paramount to the overall safety of the organization's data.

---METADATA---
VENDOR: Pandora FMS
PRODUCT: Pandora FMS
AFFECTED_VERSIONS: 777 through 800
CONFIDENCE: high
MISSING: none
---END_METADATA---

Description Summary:
A Server-Side Request Forgery (SSRF) vulnerability in the Pandora FMS API Checker extension allows an authenticated user to perform privilege escalation.

Executive Summary:
An authenticated privilege escalation vulnerability in Pandora FMS via the API Checker extension poses a significant risk to internal system integrity.

Vulnerability Details

CVE-ID: CVE-2026-30810

Affected Software: Pandora FMS

Affected Versions: 777 through 800

Vulnerability: This vulnerability is a Server-Side Request Forgery (SSRF) flaw located within the API Checker extension. The issue requires the attacker to hold low-level privileges (PR:L) to successfully execute the attack, which subsequently leads to unauthorized privilege escalation.

Business Impact

Successful exploitation allows an attacker to manipulate server-side requests, potentially accessing internal resources or escalating their current privilege level. With a CVSS score of 8.8, this flaw represents a high-severity risk that could lead to unauthorized system control and the compromise of sensitive monitoring data.

Remediation Plan

Immediate Action: Update Pandora FMS to version 802 or 800.2 immediately to resolve the SSRF flaw.

Proactive Monitoring: Monitor system logs for unusual outbound requests originating from the API Checker extension or unexpected changes in user privilege assignments.

Compensating Controls: Implement strict network egress filtering to prevent the server from reaching unauthorized internal or external endpoints until the patch is applied.

Exploitation Status

Public Exploit Available: Unknown.

Analyst Notes: As of May 14, 2026, there is no public information indicating active exploitation or a public proof-of-concept for this vulnerability. The vulnerability's reliance on authenticated access limits the initial attack surface, but its ability to escalate privileges makes it a high priority for remediation.

Analyst Recommendation

Given the potential for privilege escalation, organizations using Pandora FMS should prioritize patching to the latest versions (v802 or v800.2). Failure to address this vulnerability increases the risk of an internal attacker or compromised user account gaining elevated control over the monitoring infrastructure.

---METADATA---
VENDOR: Pandora FMS
PRODUCT: Pandora FMS
AFFECTED_VERSIONS: 777 through 800
CONFIDENCE: high
MISSING: none
---END_METADATA---

Description Summary:
A session fixation vulnerability in Pandora FMS allows a remote, unauthenticated attacker to hijack user sessions by providing a crafted session identifier.

Executive Summary:
A session fixation flaw in Pandora FMS enables remote session hijacking, potentially granting an attacker full access to user accounts.

Vulnerability Details

CVE-ID: CVE-2026-30808

Affected Software: Pandora FMS

Affected Versions: 777 through 800

Vulnerability: This is a session fixation vulnerability (CWE-384) that allows an unauthenticated remote attacker to manipulate session IDs. The attack requires user interaction (UI:P) to succeed, typically by tricking a victim into using a pre-determined session identifier.

Business Impact

Successful exploitation allows an attacker to hijack active user sessions, leading to unauthorized access to the application as the victim. Given the CVSS score of 8.1, the potential for complete account takeover and subsequent data exposure or administrative action makes this a high-severity vulnerability.

Remediation Plan

Immediate Action: Apply the vendor-supplied security updates by upgrading to version 802 or 800.2.

Proactive Monitoring: Review application access logs for irregular session patterns or multiple logins from different source IPs associated with the same session ID.

Compensating Controls: Enforce strict session management policies, including the regeneration of session tokens upon every successful login, to mitigate the risk of fixation.

Exploitation Status

Public Exploit Available: Unknown.

Analyst Notes: As of May 14, 2026, there is no public information indicating active exploitation or a public proof-of-concept for this vulnerability. The flaw is inherently dangerous as it bypasses standard authentication mechanisms by pre-fixing the session state.

Analyst Recommendation

This vulnerability requires immediate attention due to the ease with which an attacker can compromise user sessions. IT administrators must ensure that all instances of Pandora FMS are updated to the patched versions (v802 or v800.2) to prevent unauthorized account access.

---METADATA---
VENDOR: Pandora FMS
PRODUCT: Pandora FMS
AFFECTED_VERSIONS: 777 through 800
CONFIDENCE: high
MISSING: none
---END_METADATA---

Description Summary:
A Cross-Site Request Forgery (CSRF) vulnerability in Pandora FMS allows an unauthenticated attacker to force users to perform unauthorized actions via a crafted web page.

Executive Summary:
A CSRF vulnerability in Pandora FMS allows remote attackers to perform unauthorized actions on behalf of authenticated users, risking system configuration integrity.

Vulnerability Details

CVE-ID: CVE-2026-30807

Affected Software: Pandora FMS

Affected Versions: 777 through 800

Vulnerability: This is a Cross-Site Request Forgery (CWE-352) vulnerability that allows an attacker to execute unauthorized commands or changes. The attack is unauthenticated but requires the victim to interact with a crafted malicious web page (UI:P).

Business Impact

Exploitation allows attackers to perform actions with the privileges of the victim, which could lead to unauthorized configuration changes or administrative actions within the Pandora FMS environment. The CVSS score of 8.8 reflects the high potential for impact on system integrity and unauthorized state changes.

Remediation Plan

Immediate Action: Update Pandora FMS to the patched versions v802 or v800.2 immediately.

Proactive Monitoring: Monitor application logs for unexpected administrative actions performed by users who did not initiate them.

Compensating Controls: Deploy a Web Application Firewall (WAF) with CSRF protection rules to inspect and block malicious requests that lack proper anti-forgery tokens.

Exploitation Status

Public Exploit Available: Unknown.

Analyst Notes: As of May 14, 2026, there is no public information indicating active exploitation or a public proof-of-concept for this vulnerability. The risk is driven by the ability to trick an authenticated administrator into performing actions they did not intend, potentially causing significant operational disruption.

Analyst Recommendation

Organizations should prioritize the update to version 802 or 800.2 to neutralize this CSRF vector. Ensuring that administrative sessions are protected against unauthorized requests is critical for maintaining the security posture of the Pandora FMS deployment.

---METADATA---
VENDOR: Vivotek
PRODUCT: FD8136 Camera
AFFECTED_VERSIONS: Vivotek FD8136 cameras running firmware version FD8136-VVTK-0300a
---END_METADATA---

Description Summary:
A remote buffer overflow vulnerability in the Vivotek FD8136 camera firmware allows authenticated attackers to execute arbitrary code as root.

Executive Summary:
A high-severity remote buffer overflow in Vivotek FD8136 cameras enables authenticated attackers to gain full root control over the hardware.

Vulnerability Details

CVE-ID: CVE-2026-30652

Affected Software: Vivotek FD8136

Affected Versions: Firmware version FD8136-VVTK-0300a

Vulnerability: The vulnerability resides in the /cgi-bin/dido/setdo.cgi endpoint. An authenticated attacker can trigger this buffer overflow to execute arbitrary commands with elevated system privileges.

Business Impact

The CVSS score of 8.8 highlights the significant risk of device compromise. Successful exploitation grants the attacker full root access, which can be utilized to disable security monitoring, exfiltrate video feeds, or launch further attacks against the local network.

Remediation Plan

Immediate Action: Update the camera firmware to the latest secure version released by Vivotek.

Proactive Monitoring: Review system logs for signs of unauthorized requests to the /cgi-bin/dido/setdo.cgi endpoint and monitor for unexpected process execution on the device.

Compensating Controls: Implement network-level access controls to ensure that only trusted administrative workstations can communicate with the camera's management interface.

Exploitation Status

Public Exploit Available: True

Analyst Notes: As of June 4, 2026, public exploit information exists. The capability to achieve root execution via this endpoint necessitates immediate firmware remediation.

Analyst Recommendation

This vulnerability is highly dangerous due to the potential for complete device takeover. Organizations must prioritize firmware updates for all affected Vivotek cameras and ensure that camera management interfaces are isolated from broader network segments to minimize the attack surface.

---METADATA---
VENDOR: Vivotek
PRODUCT: FD8136 Camera
AFFECTED_VERSIONS: Vivotek FD8136 cameras running firmware version FD8136-VVTK-0300a
---END_METADATA---

Description Summary:
A post-authentication remote buffer overflow in the Vivotek FD8136 camera firmware allows authenticated attackers to execute arbitrary code as root.

Executive Summary:
A critical post-authentication remote buffer overflow in Vivotek FD8136 cameras allows an attacker to achieve full root-level control over the device.

Vulnerability Details

CVE-ID: CVE-2026-30650

Affected Software: Vivotek FD8136

Affected Versions: Firmware version FD8136-VVTK-0300a

Vulnerability: This buffer overflow exists in the /cgi-bin/admin/eventtask.cgi endpoint. It requires an authenticated user to trigger, subsequently allowing the execution of arbitrary code with root privileges.

Business Impact

With a CVSS score of 8.8, this vulnerability represents a severe threat to network security. Compromised cameras can be used as a pivot point for lateral movement within the network, or to facilitate botnet recruitment, surveillance interception, and persistent unauthorized access to the facility's physical security infrastructure.

Remediation Plan

Immediate Action: Apply the latest firmware update provided by Vivotek for the FD8136 model to patch the buffer overflow.

Proactive Monitoring: Monitor network traffic for unusual outbound connections from camera devices and inspect access logs for repeated failed attempts to access administrative CGI endpoints.

Compensating Controls: Isolate camera management interfaces on a dedicated, non-routable VLAN and restrict access to these interfaces using strict firewall rules.

Exploitation Status

Public Exploit Available: True

Analyst Notes: As of June 4, 2026, public exploit information is available. The ability to achieve root-level execution makes this a critical risk for any environment relying on these devices.

Analyst Recommendation

Due to the availability of exploits and the severity of root-level command execution, immediate firmware patching is required. Administrators should ensure these devices are not exposed to the public internet and restrict management access to authorized personnel only.

---METADATA---
VENDOR: DedeCMS
PRODUCT: DedeCMS
AFFECTED_VERSIONS: 5.7.118
---END_METADATA---

Description Summary:
A Remote Code Execution (RCE) vulnerability exists in DedeCMS 5.7.118. Attackers can execute arbitrary code by utilizing crafted setup tag values during a module upload process.

Executive Summary:
A critical remote code execution vulnerability in DedeCMS allows attackers to gain full system control by uploading malicious module setup tags.

Vulnerability Details

CVE-ID: CVE-2026-30643

Affected Software: DedeCMS

Affected Versions: 5.7.118

Vulnerability: This vulnerability is a Remote Code Execution (RCE) flaw residing in the module upload functionality. An attacker can bypass security restrictions by injecting malicious code into setup tag values, which the server subsequently executes during processing.

Business Impact

A successful exploit allows for complete server compromise, enabling attackers to steal sensitive data, modify website content, or deploy ransomware. Given the CVSS score of 9.8, this vulnerability poses a catastrophic risk to the confidentiality, integrity, and availability of the affected environment.

Remediation Plan

Immediate Action: Administrators should immediately update DedeCMS to the latest available security patch or version that addresses this flaw.

Proactive Monitoring: Monitor server logs for unusual module upload activity and inspect the filesystem for unauthorized PHP files or modified setup configuration files.

Compensating Controls: Implement strict IP whitelisting for the administrative backend and utilize a Web Application Firewall (WAF) to block suspicious upload requests containing PHP tags or shell commands.

Exploitation Status

Public Exploit Available: false

Analyst Notes: As of Apr 1, 2026, there is no public information indicating active exploitation of this vulnerability. However, because RCE via file upload is a common target for automated exploit kits, the potential for rapid exploitation is extremely high.

Analyst Recommendation

The critical nature of this RCE vulnerability necessitates immediate remediation. Organizations using DedeCMS 5.7.118 must prioritize patching this software today to prevent total system takeover and subsequent data exfiltration.

---METADATA---
VENDOR: automagik
PRODUCT: automagik-genie
AFFECTED_VERSIONS: See vendor advisory
CONFIDENCE: low
MISSING: versions, patch, technical_details
---END_METADATA---

Description Summary:
A command injection vulnerability has been identified in the automagik-genie software, potentially allowing remote attackers to execute arbitrary commands on the underlying system.

Executive Summary:
The automagik-genie utility contains a command injection vulnerability that could allow unauthenticated remote attackers to achieve remote code execution.

Vulnerability Details

CVE-ID: CVE-2026-30635

Affected Software: automagik-genie

Affected Versions: See vendor advisory

Vulnerability: The software is susceptible to command injection, where an attacker can supply malicious input that is executed by the host system. This vulnerability does not require authentication (PR:N) and carries a significant risk of full system compromise (AV:N/AC:H/PR:N/UI:N).

Business Impact

Command injection is a critical vulnerability that typically leads to full system compromise, allowing an attacker to steal data, install malware, or pivot into the internal network. With a CVSS score of 8.1, the potential for catastrophic business impact, including total loss of system integrity and confidentiality, is extremely high.

Remediation Plan

Immediate Action: Check the official automagik repository or vendor advisory for security patches and apply them immediately.

Proactive Monitoring: Monitor system logs for suspicious command-line arguments or unexpected child processes being spawned by the automagik-genie application.

Compensating Controls: Deploy a Web Application Firewall (WAF) or an Intrusion Detection System (IDS) with rules configured to detect and block common shell metacharacters and command injection payloads.

Exploitation Status

Public Exploit Available: Unknown.

Analyst Notes: As of May 13, 2026, there is no confirmed active exploitation in the wild; however, per CISA's SSVC assessment a proof-of-concept exists, so exploitation risk should be treated as credible. The complexity of the attack is listed as high, which may serve as a temporary barrier for less sophisticated actors.

Analyst Recommendation

Due to the existence of a proof-of-concept and the high severity of command injection, administrators must prioritize identifying instances of automagik-genie within their environment. If a vendor patch is unavailable, consider restricting network access to the application or disabling the vulnerable component until a secure version is released.

---METADATA---
VENDOR: knowns-dev
PRODUCT: knowns
AFFECTED_VERSIONS: knowns-dev/knowns < 0.11.4
CONFIDENCE: high
MISSING: none
---END_METADATA---

Description Summary:
A directory traversal vulnerability in knowns-dev/knowns allows unauthenticated attackers to access unauthorized files via crafted path values in the get_doc and update_doc tools.

Executive Summary:
A directory traversal vulnerability in knowns-dev/knowns allows unauthenticated remote attackers to read sensitive files, posing a significant security risk.

Vulnerability Details

CVE-ID: CVE-2026-30633

Affected Software: knowns-dev knowns

Affected Versions: knowns-dev/knowns < 0.11.4

Vulnerability: This is a directory traversal vulnerability occurring within the get_doc and update_doc tools. An unauthenticated attacker can supply a crafted path value to bypass security controls and access unauthorized portions of the file system.

Business Impact

Successful exploitation allows unauthorized access to sensitive data stored on the host system. Given the CVSS score of 7.5, this high-severity vulnerability could lead to the exposure of configuration files, credentials, or proprietary information, directly impacting data confidentiality and potentially facilitating further attacks.

Remediation Plan

Immediate Action: Update knowns-dev/knowns to version 0.11.4 or later, which implements case-insensitive matching and per-segment canonicalization to neutralize traversal attempts.

Proactive Monitoring: Review web server and application logs for suspicious path traversal patterns, such as sequences containing "../" or encoded directory navigation characters.

Compensating Controls: Deploy a Web Application Firewall (WAF) with rules configured to detect and block directory traversal attempts targeting application parameters.

Exploitation Status

Public Exploit Available: Yes, a public proof-of-concept exists as indicated by the available data.

Analyst Notes: As of July 20, 2026, there is no confirmed active exploitation in the wild; however, per CISA's SSVC assessment a proof-of-concept exists, so exploitation risk should be treated as credible. The flaw is inherently dangerous due to its low attack complexity and lack of required authentication.

Analyst Recommendation

This vulnerability presents a high risk to organizational data integrity and confidentiality. Administrators must prioritize upgrading to version 0.11.4 immediately to resolve the directory traversal flaw and prevent potential unauthorized file system access.

---METADATA---
VENDOR: bytebot-ai
PRODUCT: bytebot-ai
AFFECTED_VERSIONS: prior to commit 3d37894ce07ef8d8b40adc7fd309ad96c2a71313
CONFIDENCE: high
MISSING: none
---END_METADATA---

Description Summary:
A vulnerability in bytebot-ai allows an unauthenticated attacker to execute arbitrary code via a crafted path provided to the computer_write_file function.

Executive Summary:
A critical remote code execution vulnerability in bytebot-ai allows unauthenticated attackers to write arbitrary files and execute code on the host system.

Vulnerability Details

CVE-ID: CVE-2026-30631

Affected Software: bytebot-ai

Affected Versions: prior to commit 3d37894ce07ef8d8b40adc7fd309ad96c2a71313

Vulnerability: The application fails to properly sanitize file paths passed to the computer_write_file function. This allows an unauthenticated attacker to write files to arbitrary locations, leading to arbitrary code execution on the underlying system.

Business Impact

With a critical CVSS score of 9.8, this vulnerability represents a severe risk of full system compromise. An attacker could gain persistent access, exfiltrate sensitive data, or disrupt business operations by overwriting critical system files, making immediate remediation essential for any environment running this software.

Remediation Plan

Immediate Action: Update the bytebot-ai software to at least commit 3d37894ce07ef8d8b40adc7fd309ad96c2a71313.

Proactive Monitoring: Monitor file system activity, specifically tracking writes to sensitive directories or unexpected file creation events initiated by the bytebot-ai process.

Compensating Controls: Restrict the service account permissions running bytebot-ai to the absolute minimum required, and implement file system integrity monitoring to detect unauthorized changes.

Exploitation Status

Public Exploit Available: No (no confirmed public exploit)

Analyst Notes: As of July 20, 2026, there is no confirmed active exploitation in the wild; however, per CISA's SSVC assessment, a proof-of-concept exists, so exploitation risk should be treated as credible. The flaw is inherently dangerous due to its ability to facilitate unauthenticated file writes.

Analyst Recommendation

Given the critical nature of this vulnerability and the existence of a proof-of-concept, administrators should apply the fix immediately. The potential for unauthenticated remote code execution makes this a high-priority target for malicious actors.

---METADATA---
VENDOR: BerriAI
PRODUCT: LiteLLM
AFFECTED_VERSIONS: Prior to v1.83.6-nightly / v1.83.7-stable
CONFIDENCE: high
MISSING: none
---END_METADATA---

Description Summary:
LiteLLM versions prior to v1.83.6-nightly and v1.83.7-stable are vulnerable to remote code execution due to improper validation of command configurations for MCP servers.

Executive Summary:
An authenticated remote code execution vulnerability in LiteLLM allows users with the PROXY_ADMIN role to execute arbitrary commands by injecting malicious JSON configurations.

Vulnerability Details

CVE-ID: CVE-2026-30623

Affected Software: BerriAI LiteLLM

Affected Versions: Prior to v1.83.6-nightly / v1.83.7-stable

Vulnerability: The vulnerability exists in the MCP server creation functionality, where the application fails to validate arbitrary command and argument values provided in JSON configurations. Exploitation requires the attacker to possess a valid API key and hold the PROXY_ADMIN role, allowing them to execute code on the host.

Business Impact

The CVSS score of 9.8 highlights the critical impact of this vulnerability, as it allows an attacker with elevated privileges to achieve remote code execution. This could lead to full system takeover, unauthorized access to sensitive API keys, and potential exfiltration of data processed by the LiteLLM proxy.

Remediation Plan

Immediate Action: Update LiteLLM to version v1.83.6-nightly or v1.83.7-stable immediately to resolve the command injection vulnerability.

Proactive Monitoring: Audit existing LiteLLM configurations for any suspicious MCP server entries and review logs for commands executed via the proxy admin interface.

Compensating Controls: Apply the principle of least privilege by strictly limiting the number of users assigned the PROXY_ADMIN role and ensuring the application is hosted in a restricted environment.

Exploitation Status

Public Exploit Available: No

Analyst Notes: As of July 19, 2026, there is no public information indicating active exploitation or a public proof-of-concept for this vulnerability. This flaw represents a significant risk within the context of the Model Context Protocol ecosystem.

Analyst Recommendation

Administrators should prioritize upgrading to the patched versions of LiteLLM to eliminate this command injection risk. Furthermore, strictly control access to administrative roles to prevent unauthorized users from leveraging this functionality for malicious purposes.

---METADATA---
VENDOR: xszyou
PRODUCT: Fay
AFFECTED_VERSIONS: 4.3.1
CONFIDENCE: high
MISSING: patch
---END_METADATA---

Description Summary:
The Fay framework version 4.3.1 contains a remote code execution vulnerability in its MCP STDIO server management interface, allowing unauthenticated attackers to execute arbitrary commands.

Executive Summary:
A critical remote code execution vulnerability in Fay 4.3.1 allows unauthenticated attackers to manipulate the MCP management interface to execute arbitrary system commands.

Vulnerability Details

CVE-ID: CVE-2026-30618

Affected Software: xszyou Fay

Affected Versions: 4.3.1

Vulnerability: The application exposes an MCP management interface that lacks proper authentication and input validation. This allows a remote, unauthenticated attacker to configure an MCP STDIO server with arbitrary commands and parameters, which the system subsequently executes on the host.

Business Impact

With a CVSS score of 9.8, this vulnerability poses an extreme risk to the integrity and availability of the affected system. An attacker can gain full control over the server, leading to potential data theft, lateral movement within the network, and the total compromise of the host environment.

Remediation Plan

Immediate Action: Immediately isolate the Fay management interface from the public internet and evaluate the vendor advisory for any available security updates or configuration hardening steps.

Proactive Monitoring: Review access logs for suspicious requests directed at the MCP management endpoints and monitor for any unauthorized child processes spawned by the main application.

Compensating Controls: Implement network-level access control lists (ACLs) to limit access to the management interface to trusted administrative IP addresses only.

Exploitation Status

Public Exploit Available: No

Analyst Notes: As of July 19, 2026, there is no public information indicating active exploitation or a public proof-of-concept for this vulnerability. This issue stems from a broader systemic weakness in the current implementation of the Model Context Protocol.

Analyst Recommendation

The severity of this RCE vulnerability necessitates immediate defensive action. Organizations must ensure that the Fay management interface is not exposed to the public internet and should monitor for any signs of unauthorized configuration changes until a formal patch is deployed.

---METADATA---
VENDOR: SGLang
PRODUCT: SGLang
AFFECTED_VERSIONS: See vendor advisory
---END_METADATA---

Description Summary:
SGLang's encoder parallel disaggregation system allows unauthenticated remote code execution through insecure deserialization in the disaggregation module.

Executive Summary:
An unauthenticated remote code execution vulnerability exists in the SGLang disaggregation module, posing a critical risk to AI processing environments.

Vulnerability Details

CVE-ID: CVE-2026-3060

Affected Software: SGLang

Affected Versions: See vendor advisory for specific affected versions

Vulnerability: The disaggregation module in the SGLang encoder parallel system uses pickle.loads() on untrusted data without any authentication. This allows a remote attacker to execute arbitrary commands by sending a specially crafted pickle payload.

Business Impact

This flaw carries a CVSS score of 9.8, indicating a nearly maximum level of risk. An attacker can gain full control over the AI encoder systems, leading to the compromise of proprietary models, sensitive training data, and the underlying server infrastructure, resulting in significant operational disruption.

Remediation Plan

Immediate Action: Apply the vendor-provided update to SGLang immediately to patch the insecure disaggregation module.

Proactive Monitoring: Review system logs for Python-related execution errors or unexpected outbound connections from the SGLang service.

Compensating Controls: Utilize a Web Application Firewall (WAF) or deep packet inspection to block unauthorized ZMQ or disaggregation-related traffic at the perimeter.

Exploitation Status

Public Exploit Available: No

Analyst Notes: As of Mar 12, 2026, there is no public information indicating active exploitation of this vulnerability. The technical nature of the flaw—unauthenticated insecure deserialization—makes it a high-priority target for automated scanning and exploitation.

Analyst Recommendation

The unauthenticated nature of this RCE makes it an urgent security concern. We strongly recommend that all users of SGLang update their installations immediately and ensure that disaggregation modules are not exposed to untrusted networks.

---METADATA---
VENDOR: SGLang
PRODUCT: SGLang
AFFECTED_VERSIONS: See vendor advisory
---END_METADATA---

Description Summary:
SGLang's multimodal generation module is vulnerable to unauthenticated remote code execution via untrusted data deserialization in the ZMQ broker.

Executive Summary:
The SGLang multimodal generation module is susceptible to a critical unauthenticated remote code execution vulnerability due to insecure deserialization of untrusted data.

Vulnerability Details

CVE-ID: CVE-2026-3059

Affected Software: SGLang

Affected Versions: See vendor advisory for specific affected versions

Vulnerability: This vulnerability occurs in the ZMQ broker of the multimodal generation module, which uses pickle.loads() to deserialize data without authentication. An unauthenticated remote attacker can send a crafted payload to trigger arbitrary code execution.

Business Impact

A successful exploit allows a completely unauthenticated attacker to execute code on the server hosting the SGLang module. Given the CVSS score of 9.8, this poses a severe risk to AI infrastructure, potentially leading to data theft, model manipulation, or the use of the server for further attacks into the corporate network.

Remediation Plan

Immediate Action: Update SGLang to the latest version that replaces insecure deserialization methods with secure alternatives.

Proactive Monitoring: Monitor network traffic for unusual ZMQ protocol activity and inspect logs for unauthorized connections to the generation module ports.

Compensating Controls: Implement network-level access controls to ensure the ZMQ broker is only accessible from trusted internal IP addresses and is not exposed to the public internet.

Exploitation Status

Public Exploit Available: No

Analyst Notes: As of Mar 12, 2026, there is no public information indicating active exploitation of this vulnerability. However, the use of pickle.loads() is a well-known insecure pattern, making the development of an exploit relatively straightforward for attackers.

Analyst Recommendation

This is a critical vulnerability that must be addressed immediately. Organizations using SGLang for multimodal AI generation should apply the latest security updates and restrict network access to the vulnerable module to prevent unauthenticated remote exploitation.

---METADATA---
VENDOR: SourceCodester
PRODUCT: Sales and Inventory System
AFFECTED_VERSIONS: 1.0
---END_METADATA---

Description Summary:
SourceCodester Sales and Inventory System 1.0 is vulnerable to Reflected Cross-Site Scripting (XSS) via the 'msg' parameter in add_stock.php due to insufficient input sanitization.

Executive Summary:
A critical Reflected Cross-Site Scripting (XSS) vulnerability in SourceCodester Sales and Inventory System allows unauthenticated attackers to execute malicious scripts in the context of a user's browser session.

Vulnerability Details

CVE-ID: CVE-2026-30562

Affected Software: SourceCodester Sales and Inventory System

Affected Versions: 1.0

Vulnerability: The application fails to sanitize the "msg" parameter within the add_stock.php file. This allows a remote, unauthenticated attacker to craft a malicious URL containing arbitrary web scripts or HTML, which is then executed by the victim's browser when the link is accessed.

Business Impact

While XSS is often categorized as medium severity, the assigned CVSS score of 9.3 indicates a critical risk level in this specific environment, likely due to the potential for administrative session hijacking. Attackers can steal session cookies, perform unauthorized actions on behalf of users, or redirect victims to malicious phishing sites, leading to significant data loss and reputational damage.

Remediation Plan

Immediate Action: Apply the latest vendor patches immediately or manually implement input encoding/sanitization on the "msg" parameter in add_stock.php to prevent script execution.

Proactive Monitoring: Monitor web application firewall (WAF) logs for common XSS patterns, such as script tags or encoded JavaScript signatures, targeting the add_stock.php endpoint.

Compensating Controls: Deploy a Content Security Policy (CSP) header to restrict the execution of unauthorized inline scripts and prevent the loading of scripts from untrusted external domains.

Exploitation Status

Public Exploit Available: false

Analyst Notes: As of Mar 30, 2026, there is no public information indicating active exploitation of this vulnerability. However, Reflected XSS is a common target for social engineering and phishing campaigns, making the potential for exploitation high.

Analyst Recommendation

Given the critical CVSS score of 9.3, this vulnerability must be addressed with high urgency. Organizations should prioritize sanitizing all user-controllable inputs and consider moving to a more secure, actively maintained sales and inventory platform if vendor support is unavailable.

---METADATA---
VENDOR: SourceCodester
PRODUCT: Online Food Ordering System
AFFECTED_VERSIONS: v1.0
---END_METADATA---

Description Summary:
A SQL Injection vulnerability exists in SourceCodester Online Food Ordering System v1.0 within the admin/manage_product.php file via the "id" parameter.

Executive Summary:
A critical SQL injection vulnerability in SourceCodester Online Food Ordering System v1.0 allows for unauthorized database manipulation through the administrative product management interface.

Vulnerability Details

CVE-ID: CVE-2026-30533

Affected Software: SourceCodester Online Food Ordering System

Affected Versions: v1.0

Vulnerability: The admin/manage_product.php file fails to properly sanitize the id parameter, leading to a SQL Injection vulnerability. This flaw resides within the administrative module, potentially allowing an attacker to bypass authentication or escalate privileges.

Business Impact

The impact of this vulnerability is severe, as reflected by its CVSS score of 9.8. An attacker can gain unauthorized access to the database, leading to the theft of customer records or the alteration of financial and product data. Such a breach would likely result in system downtime and a loss of customer trust in the platform's security.

Remediation Plan

Immediate Action: Immediately update the Online Food Ordering System to the latest version or manually implement input validation and parameterized queries for the id parameter in manage_product.php.

Proactive Monitoring: Review database logs for unauthorized query executions and audit administrative account activity for signs of hijacking or credential misuse.

Compensating Controls: Utilize a Web Application Firewall (WAF) to inspect and block malicious payloads containing SQL syntax targeting the administrative subdirectories.

Exploitation Status

Public Exploit Available: false

Analyst Notes: As of Mar 27, 2026, there is no public information indicating active exploitation of this vulnerability. However, because the vulnerability is located in a common management script, it is a likely target for automated vulnerability scanners.

Analyst Recommendation

The severity of a 9.8 CVSS score necessitates immediate remediation. Organizations using this software must apply security updates or move the administrative interface behind a strictly controlled VPN or IP-allowlist to mitigate the risk of remote exploitation.

---METADATA---
VENDOR: SourceCodester
PRODUCT: Online Food Ordering System
AFFECTED_VERSIONS: v1.0
---END_METADATA---

Description Summary:
A SQL Injection vulnerability exists in SourceCodester Online Food Ordering System v1.0 within the admin/view_product.php file via the "id" parameter.

Executive Summary:
SourceCodester Online Food Ordering System v1.0 is vulnerable to a critical SQL injection flaw that could allow an attacker to compromise the underlying database.

Vulnerability Details

CVE-ID: CVE-2026-30532

Affected Software: SourceCodester Online Food Ordering System

Affected Versions: v1.0

Vulnerability: This vulnerability is a SQL Injection located in the admin/view_product.php file. An attacker can manipulate the id parameter to execute arbitrary SQL commands against the database, potentially targeting administrative interfaces.

Business Impact

Successful exploitation of this vulnerability could lead to total loss of confidentiality, integrity, and availability of the application's data. Given the CVSS score of 9.8, an attacker could extract sensitive user information, modify menu data, or gain full administrative control over the food ordering platform, resulting in significant operational disruption and reputational damage.

Remediation Plan

Immediate Action: Administrators should immediately update the software to the latest available version or apply custom patches to sanitize the id parameter using prepared statements.

Proactive Monitoring: Monitor web server access logs for unusual characters or SQL keywords (e.g., SELECT, UNION, SLEEP) within the URL parameters of the admin/ directory.

Compensating Controls: Deploy a Web Application Firewall (WAF) with active SQL injection protection rules to filter malicious traffic targeting the vulnerable PHP scripts.

Exploitation Status

Public Exploit Available: false

Analyst Notes: As of Mar 27, 2026, there is no public information indicating active exploitation of this vulnerability. However, due to the nature of the flaw and the ease of automating SQL injection attacks, the potential for exploitation is high.

Analyst Recommendation

This vulnerability represents a critical risk to the data integrity of the Online Food Ordering System. It is highly recommended that the application be taken offline or patched immediately to prevent unauthorized database access. Prioritize the implementation of parameterized queries to eliminate the root cause of the injection.

---METADATA---
VENDOR: SourceCodester
PRODUCT: Online Food Ordering System
AFFECTED_VERSIONS: v1.0
---END_METADATA---

Description Summary:
SourceCodester Online Food Ordering System v1.0 contains a critical SQL injection vulnerability in the save_customer action within Actions.php.

Executive Summary:
Unauthenticated attackers can execute arbitrary SQL commands on the Online Food Ordering System database, leading to the full compromise of customer data and system credentials.

Vulnerability Details

CVE-ID: CVE-2026-30530

Affected Software: SourceCodester Online Food Ordering System

Affected Versions: v1.0

Vulnerability: A SQL injection vulnerability exists in the Actions.php file. The application fails to properly sanitize user input provided to the "username" parameter during the save_customer action. This allows an unauthenticated attacker to inject malicious SQL commands directly into the database query.

Business Impact

Successful exploitation allows for the complete exfiltration of the customer database, including personally identifiable information (PII) and hashed passwords. Attackers may also be able to modify database records or gain administrative access to the application. The CVSS score of 9.8 indicates a critical risk to data integrity and confidentiality.

Remediation Plan

Immediate Action: Manually patch the Actions.php file to use prepared statements or parameterized queries for the save_customer action.

Proactive Monitoring: Review web server and database logs for common SQL injection patterns, such as single quotes, comments (--), or UNION SELECT statements in POST requests.

Compensating Controls: Implement a Web Application Firewall (WAF) to filter malicious input to the Actions.php endpoint.

Exploitation Status

Public Exploit Available: false

Analyst Notes: As of Mar 27, 2026, there is no public information indicating active exploitation. However, SQL injection in open-source scripts is frequently targeted by automated scanners.

Analyst Recommendation

Users of this system must apply manual code fixes immediately, as v1.0 is highly vulnerable to automated attacks. Transitioning to a more secure, actively maintained platform is strongly recommended for production environments handling sensitive customer information.