13 vulnerabilities across 10 products scored HIGH or above on September 23, 2026.

  • 🟠 HIGH: 13

Exploit Status Upgrades

The following CVEs from previous bulletins have been upgraded based on new exploit intelligence:

  • [UPGRADED] CVE-2026-28325 (solarwinds/observability_self-hosted) — F1: exploitable → functional, AAS: 9.6 → 11.6 (HIGH → HIGH). Originally in 2026-09-22 bulletin.
  • [UPGRADED] CVE-2026-77521 (1panel-dev/maxkb) — F1: exploitable → functional, AAS: 11.2 → 13.2 (HIGH → CRITICAL). Originally in 2026-09-21 bulletin.
  • [UPGRADED] CVE-2026-93603 (patriksimek/vm2) — F1: exploitable → functional, AAS: 12.7 → 14.7 (CRITICAL → CRITICAL). Originally in 2026-09-18 bulletin.
  • [UPGRADED] CVE-2026-93605 (patriksimek/vm2) — F1: exploitable → functional, AAS: 12.7 → 14.7 (CRITICAL → CRITICAL). Originally in 2026-09-18 bulletin.
  • [UPGRADED] CVE-2026-93606 (patriksimek/vm2) — F1: theoretical → functional, AAS: 11.7 → 14.7 (HIGH → CRITICAL). Originally in 2026-09-18 bulletin.
  • [UPGRADED] CVE-2026-81657 (ibm/guardium_data_protection) — F1: exploitable → functional, AAS: 10.9 → 12.9 (HIGH → CRITICAL). Originally in 2026-09-18 bulletin.
  • [UPGRADED] CVE-2026-84383 (strukturag/libheif) — F1: exploitable → functional, AAS: 10.9 → 12.9 (HIGH → CRITICAL). Originally in 2026-09-18 bulletin.
  • [UPGRADED] CVE-2026-12793 (jetmonsters/jetformbuilder_—_dynamic_blocks_form_builder) — F1: exploitable → functional, AAS: 10.1 → 12.1 (HIGH → CRITICAL). Originally in 2026-09-16 bulletin.

🟠 [HIGH] rabbitmq/rabbitmq-server

2 CVEs | CVSS 4.0: 9.2 | AAS 11.2

  • cpe:2.3:a:rabbitmq:rabbitmq-server:*:*:*:*:*:*:*:* (< 3.13.15)
  • cpe:2.3:a:rabbitmq:rabbitmq-server:*:*:*:*:*:*:*:* (>= 4.0.0, < 4.0.20)
  • cpe:2.3:a:rabbitmq:rabbitmq-server:*:*:*:*:*:*:*:* (>= 4.1.0, < 4.1.11)
  • cpe:2.3:a:rabbitmq:rabbitmq-server:*:*:*:*:*:*:*:* (>= 4.2.0, < 4.2.6)

RabbitMQ Server is affected by two high-severity vulnerabilities, including at least one that allows a man-in-the-middle attacker to forge JWKS responses and force the broker to accept arbitrary JWTs when the OAuth2 plugin is in use without a properly configured CA bundle. This is particularly dangerous in containerized deployments where minimal base images may lack OS CA bundles, leaving TLS verification silently disabled. The maximum CVSS 4.0 score across these vulnerabilities is 9.2.

Organizations running RabbitMQ with OAuth2 authentication enabled should treat this as urgent. Upgrade immediately to versions 3.13.15, 4.0.20, 4.1.11, 4.2.6, or 4.3.0 or later. As an interim measure, verify that a valid cacertfile is explicitly configured in your ssl_options and confirm the OS CA bundle is present and readable in all deployment environments. Refer to the vendor advisory at the linked GitHub release page for full details.

Vendor Advisory

🟠 [HIGH] red_hat/red_hat_ansible_automation_platform_2.4_for_rhel_8

3 CVEs | CVSS 3.1: 9.9 | AAS 10.1

  • cpe:2.3:a:redhat:red_hat_ansible_automation_platform_2.4_for_rhel_8:*:*:*:*:*:*:*:*

Red Hat Ansible Automation Platform 2.4 for RHEL 8 is affected by three high-severity vulnerabilities, including multiple flaws in the automation-controller that expose the provisioning-callback secret to low-privileged users and allow attackers to spoof the calling host via a trusted X-Forwarded-For header when the controller sits behind the AAP gateway with an empty proxy allow-list. The maximum CVSS 3.1 score across these vulnerabilities is 9.9, reflecting near-complete compromise potential for affected environments.

Organizations using Ansible Automation Platform 2.4 on RHEL 8, especially those with gateway-fronted deployments, should prioritize patching immediately. Apply the fixes provided in Red Hat’s advisory RHSA-2026:71113 and review job template permissions and proxy allow-list configurations to ensure provisioning-callback secrets are not accessible to read-only users.

Vendor Advisory

🟠 [HIGH] plank/laravel-mediable

1 CVE | CVSS 4.0: 9.3 | AAS 9.9

  • cpe:2.3:a:plank:laravel-mediable:*:*:*:*:*:*:*:*

Laravel-Mediable versions 7.0.0 through 7.0.1 contain a high-severity file upload bypass vulnerability caused by an incomplete patch for CVE-2026-49972, where the .pht extension was omitted from the forbidden extensions blocklist. On Debian and Ubuntu systems running Apache, the .pht extension is executed as PHP by default, allowing an attacker to upload a malicious file that bypasses all validation and achieves remote code execution. The CVSS 4.0 score for this vulnerability is 9.3.

Teams running Laravel-Mediable 7.0.0 or 7.0.1 on Apache-based Linux deployments should upgrade to version 7.0.2 immediately. As a temporary mitigation, manually add pht to the forbidden_extensions list in config/mediable.php and review any previously uploaded files for the .pht extension.

Vendor Advisory

🟠 [HIGH] mlflow/mlflow

1 CVE | CVSS 3.1: 8.8 | AAS 9.6

  • cpe:2.3:a:mlflow:mlflow:*:*:*:*:*:*:*:*

MLflow versions 2.1.0 through 3.14.0 contain a high-severity remote code execution vulnerability in the statsmodel flavor, where the _load_model() function completely ignores the MLFLOW_ALLOW_PICKLE_DESERIALIZATION=False security control. An attacker who can supply a crafted MLmodel artifact can exploit this gap to execute arbitrary code on any system that loads the malicious model, regardless of whether pickle deserialization has been explicitly disabled. The CVSS 3.1 score for this vulnerability is 8.8.

Organizations using MLflow for model serving, experimentation, or pipeline workflows should upgrade beyond version 3.14.0 as soon as a fix is available. In the interim, audit model registries and artifact stores for untrusted or unexpected MLmodel files, restrict who can publish models, and review the CERT/CC advisory at the linked URL for additional guidance.

Vendor Advisory

🟠 [HIGH] red_hat/red_hat_enterprise_linux_10

1 CVE | CVSS 3.1: 8.8 | AAS 9.6

  • cpe:2.3:a:redhat:enterprise_linux:*:*:*:*:*:*:*:*

Red Hat Enterprise Linux 10 is affected by a high-severity vulnerability in Flatpak where a malicious or compromised repository can write attacker-controlled content to arbitrary locations on the host filesystem. The flaw combines unsanitized path traversal in blob names from xa.extra-data-sources with symlink-following behavior in extract_extra_data(), and on system-level installs the write occurs as root, enabling full host compromise. The CVSS 3.1 score for this vulnerability is 8.8.

Organizations running Flatpak on RHEL 10, particularly those allowing third-party or untrusted Flatpak repositories, should apply patches from Red Hat as soon as they are available. In the meantime, restrict Flatpak installations to trusted repositories only, avoid system-level installs from unverified sources, and monitor the linked Red Hat advisory for updated remediation guidance.

Vendor Advisory

🟠 [HIGH] wbw_plugins/product_filter_by_wbw

1 CVE | CVSS 3.1: 9.3 | AAS 9.6

  • cpe:2.3:a:wbw_plugins:product_filter_by_wbw:*:*:*:*:*:*:*:*

The Product Filter by WBW plugin for WordPress versions 3.1.7 and earlier contains a high-severity unauthenticated SQL injection vulnerability, meaning no login or user interaction is required for exploitation. An attacker can remotely extract, modify, or delete database contents, potentially compromising the entire WordPress installation and any sensitive data it stores. The CVSS 3.1 score for this vulnerability is 9.3.

Any WordPress site running the Product Filter by WBW plugin should update beyond version 3.1.7 immediately or deactivate the plugin until a patched release is available. Review database logs for signs of injection attempts and consult the Patchstack advisory at the linked URL for further details.

Vendor Advisory

🟠 [HIGH] openbao/openbao

1 CVE | CVSS 4.0: 9.2 | AAS 9.6

  • cpe:2.3:a:openbao:openbao:*:*:*:*:*:*:*:*

OpenBao versions prior to 2.6.0 contain a high-severity timing side-channel vulnerability in the recovery token validation logic, where ordinary string comparison is used instead of constant-time comparison. A remote unauthenticated attacker who can send repeated recovery mode requests and measure response timing could gradually infer the highly privileged recovery token, then use it to authorize operations that read or modify secrets stored in OpenBao. The CVSS 4.0 score for this vulnerability is 9.2.

Organizations running OpenBao in any environment should upgrade to version 2.6.0 immediately. Until patching is complete, restrict network access to the recovery mode endpoint and monitor for unusual volumes of recovery-related requests that could indicate an active timing attack.

Vendor Advisory

🟠 [HIGH] tw93/pake

1 CVE | CVSS 3.1: 8.2 | AAS 9.5

  • cpe:2.3:a:tw93:pake:*:*:*:*:*:*:*:*

Pake, a tool that wraps websites into Tauri-based desktop applications, generates apps with an overly permissive default configuration that grants native IPC access to any HTTPS origin combined with globally exposed Tauri APIs. This means any untrusted web content loaded within a Pake-generated application can invoke native system functionality, potentially leading to local file access, command execution, or other host-level compromise. The CVSS 3.1 score for this vulnerability is 8.2.

Organizations or developers distributing Pake-generated desktop applications should immediately review and restrict the IPC remote URL whitelist in src-tauri/capabilities/default.json to only the intended origin, and disable withGlobalTauri unless explicitly required. Treat any previously distributed Pake-built application as potentially exploitable and rebuild with tightened configurations after consulting the vendor’s GitHub repository for updated guidance.

Vendor Advisory

🟠 [HIGH] klever-io/klever-go

1 CVE | CVSS 3.1: 8.6 | AAS 9.4

  • cpe:2.3:a:klever-io:klever-go:*:*:*:*:*:*:*:*

Klever-Go, the Go implementation of the Klever blockchain protocol, versions prior to 1.7.20 expose an unauthenticated WebSocket endpoint at /log that allows any remote client to modify process-wide logging configuration including log levels and formatting options. An attacker can exploit this to suppress security-relevant log output, flood logs to cause denial of service, or potentially extract sensitive runtime information streamed through the logging subsystem. The CVSS 3.1 score for this vulnerability is 8.6.

Node operators running Klever-Go should upgrade to version 1.7.20 or later immediately. Until patching is possible, restrict network access to the node’s API port to trusted sources only and monitor for unexpected WebSocket connections to the /log endpoint.

Vendor Advisory

🟠 [HIGH] microsoft/microsoft_365_apps_for_enterprise

1 CVE | CVSS 3.1: 8.8 | AAS 9.1

  • cpe:2.3:a:microsoft:microsoft_365_apps_for_enterprise:*:*:*:*:*:*:*:*

Microsoft 365 Apps for Enterprise is affected by a high-severity remote code execution vulnerability in Microsoft Outlook. Successful exploitation could allow an attacker to execute arbitrary code on a target system, potentially through maliciously crafted email content or attachments. The CVSS 3.1 score for this vulnerability is 8.8.

All organizations using Microsoft 365 Apps for Enterprise should apply the latest security updates from Microsoft immediately. Refer to the Microsoft Security Response Center advisory for patch availability, affected version details, and any recommended workarounds while updates are being deployed.

Vendor Advisory