Monthly
Coder deployments that run workspace agents on Google Compute Engine with GCP instance-identity authentication accept replayed GCE identity tokens because the POST /api/v2/workspaceagents/google-instance-identity endpoint never validated the token's audience claim, allowing an attacker holding an unexpired format=full token minted for a different relying party to exchange it for that workspace agent's session token. The resulting agent token stays valid until the workspace's next build, well past the identity token's roughly one-hour lifetime. Affected releases are those before 2.29.21 and the 2.30.0-2.34.11, 2.35.0-2.35.8, 2.36.0-2.36.6 and 2.37.0-2.37.3 ranges, fixed in 2.38.0 and the listed backports; the flaw is unauthenticated per the supplied vector (PR:N) but conditional, since the attacker must already possess a genuine, unexpired token from a VM backing an agent on the workspace's latest build, and no public exploit was identified at time of analysis. The vendor advisory credits independent reports from a customer security team and an external researcher.
Credential theft in multi-tenant Red Hat StreamsHub Console for Apache Kafka and Streams for Apache Kafka 2/3 deployments is possible because the operator trusts a namespace field supplied in a Console custom resource. An authenticated tenant with permission to create Console CRs (PR:L) can point spec.kafkaClusters[].credentials.kafkaUser.namespace at a victim namespace, causing the cluster-wide-privileged operator to read that namespace’s KafkaUser Secret-SASL credentials or TLS private keys-and write the plaintext into a Secret in the attacker’s namespace. CVSS 3.1 scores this 7.7 with scope changed and high confidentiality impact; no public exploit identified at time of analysis, so risk is concentrated in multi-tenant clusters where untrusted users can create Console resources.
Local privilege escalation in Windows Kernel-Mode Drivers allows an authenticated attacker who already holds low-privileged code execution on the host to elevate to a higher-integrity context, with total impact to confidentiality, integrity and availability. The CVSS vector (AV:L/AC:L/PR:L/UI:N/S:U) confirms the flaw is only reachable from a local session by a low-privileged principal, requires no user interaction, and needs no non-default feature to be enabled, so every unpatched build of Windows 10 (1607, 1809, 21H2, 22H2), Windows 11 23H2 and Windows Server 2012 through 2025 is exposed once an attacker has a foothold; there is no remote or network-reachable path. CISA SSVC records exploitation as none and no public exploit code has been identified at time of analysis; Microsoft has published a fix, making this a credible but not emergency-tier patching priority whose practical abuse case is post-access escalation on already-breached or multi-user hosts.
Authentication bypass in ZITADEL's Login V2 UI (versions 3.0.0 through 4.19.1) allows unauthenticated attackers to take over user accounts, including administrators with MFA completed, by forging unsigned session cookies and guessing time-ordered session IDs around a victim's sign-in to trigger token rotation. Exploitation requires the target to run the affected Login V2 UI and the attacker to hit a narrow timing window; no public exploit code is identified at time of analysis. While the CVE carries a CVSS 4.0 score of 9.2, the independent assessment rates it with high attack complexity (CVSS 3.1 AV:N/AC:H/PR:N/UI:N), reflecting the timing prerequisite.
Pre-authentication server-side request forgery in the SonicWall SMA1000 Appliance's Work Place interface lets a remote attacker who can reach the appliance abuse an unintended alternate access path that bypasses authentication and coerce the device into issuing requests on the attacker's behalf, reaching internal functionality that is not otherwise exposed externally (assessed CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:N - high confidentiality impact, low integrity impact, no availability impact). Affected builds are SMA1000 12.4.3-03526 (platform-hotfix) and older, and 12.5.0-02952 (platform-hotfix) and older; because the Work Place portal is normally internet-facing on these remote-access appliances, exposure is broad and this is a genuine top-priority vulnerability rather than a high-CVSS-but-low-risk artifact. No public exploit identified at time of analysis, and exploitation requires only default network reachability with no credentials and no user interaction, so the practical impact is bounded mainly by which internal services the appliance can reach through the forged requests.
Fork-originated Gitea Actions workflows can run on a base repository's registered runners without the required approval when a repository maintainer performs ordinary pull request triage, such as adding a label, because the approval decision was keyed to the user who triggered the event rather than the fork pull request author, allowing fork-controlled workflow code to execute in the base repository's CI context. Gitea versions up to and including 1.27.3 are affected when Actions is enabled and a matching runner is registered; the issue is fixed in Gitea 28.0.0. No public exploit code was identified at time of analysis and EPSS is low (0.15%, 3rd percentile), but exploitation is not autonomous: the assessed vector (CVSS:3.1/AV:N/AC:H/PR:L/UI:R/S:U/C:H/I:H/A:H) requires an attacker-submitted fork pull request plus a privileged maintainer action, and CISA SSVC rates the technical impact as total, so the outcome when the prerequisite chain is met is full compromise of confidentiality, integrity, and availability of the workflow execution environment.
Origin policy bypass in Google Chrome for Android prior to 155.0.8059.39 lets a remote attacker acting through a confused deputy have actions carried out under a different web origin's authority, so a victim who is socially engineered into opening a crafted HTML page can have that page reach resources or capabilities it should never hold for its own origin. Only Chrome on Android is affected, exploitation is not silent (it requires the user to open attacker-supplied content, CVSS PR:N/UI:R), and no public exploit code or active exploitation has been identified at time of analysis; EPSS is 0.22% (12th percentile), and Google rates the underlying Chromium bug as Low severity. Note the sharp signal conflict that must be reconciled before prioritization: the nominal vendor CVSS is 8.3 while EPSS, vendor severity, and our independent assessment (CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N) all point to materially lower real-world risk.
Confused-deputy bypass in Google Chrome's Contextual Tasks feature allows a remote attacker who already holds renderer code execution to cross a security boundary into a privileged page by loading a crafted HTML page, with high confidentiality and integrity impact (unauthenticated per the CVSS vector, but requiring user interaction). The flaw affects Chrome builds prior to 155.0.8059.39 and functions primarily as a chain link - the prerequisite renderer compromise from a separate bug is the real gate, which sharply narrows standalone risk despite the 8.1 score. No public exploit code has been identified at time of analysis and EPSS is very low at 0.18% (7th percentile), with CISA SSVC recording no observed exploitation; a vendor-released patch exists in 155.0.8059.39.
Sandbox escape and potential arbitrary code execution outside the browser sandbox in Google Chrome via a confused-deputy flaw in the Google Lens feature, affecting all desktop releases prior to 155.0.8059.39. Exploitation is not standalone: a remote attacker must first have compromised the Chrome renderer process (a separate renderer-stage bug), then convince the victim to load a crafted HTML page, after which the Lens component can be induced to perform a privileged operation that crosses the sandbox boundary on the renderer's behalf - consistent with the CVSS:3.1 vector AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H and high attack complexity. No public exploit code has been identified at time of analysis and there is no confirmed active exploitation (CISA KEV absent, SSVC exploitation: none), but the maximum technical impact rating means the issue is worth prioritizing within normal patch cycles; vendor-released patch: 155.0.8059.39.
Remote attackers can bypass system access restrictions and reach a privileged internal page in Google Chrome for Android versions prior to 155.0.8059.39 by way of a confused-deputy weakness in the SignIn component, which performs privileged navigation on behalf of an untrusted caller. Exploitation requires the victim to load attacker-controlled crafted HTML (for example, visiting a malicious or compromised page), needs no authentication or prior privileges, and is limited to the Android build on pre-patch versions. No public exploit code has been identified at time of analysis, EPSS is low at 0.22% (12th percentile), and the issue is not confirmed as actively exploited; a vendor patch (155.0.8059.39) is available, so this is a moderate-priority patch-and-move-on item rather than an emergency.
Coder deployments that run workspace agents on Google Compute Engine with GCP instance-identity authentication accept replayed GCE identity tokens because the POST /api/v2/workspaceagents/google-instance-identity endpoint never validated the token's audience claim, allowing an attacker holding an unexpired format=full token minted for a different relying party to exchange it for that workspace agent's session token. The resulting agent token stays valid until the workspace's next build, well past the identity token's roughly one-hour lifetime. Affected releases are those before 2.29.21 and the 2.30.0-2.34.11, 2.35.0-2.35.8, 2.36.0-2.36.6 and 2.37.0-2.37.3 ranges, fixed in 2.38.0 and the listed backports; the flaw is unauthenticated per the supplied vector (PR:N) but conditional, since the attacker must already possess a genuine, unexpired token from a VM backing an agent on the workspace's latest build, and no public exploit was identified at time of analysis. The vendor advisory credits independent reports from a customer security team and an external researcher.
Credential theft in multi-tenant Red Hat StreamsHub Console for Apache Kafka and Streams for Apache Kafka 2/3 deployments is possible because the operator trusts a namespace field supplied in a Console custom resource. An authenticated tenant with permission to create Console CRs (PR:L) can point spec.kafkaClusters[].credentials.kafkaUser.namespace at a victim namespace, causing the cluster-wide-privileged operator to read that namespace’s KafkaUser Secret-SASL credentials or TLS private keys-and write the plaintext into a Secret in the attacker’s namespace. CVSS 3.1 scores this 7.7 with scope changed and high confidentiality impact; no public exploit identified at time of analysis, so risk is concentrated in multi-tenant clusters where untrusted users can create Console resources.
Local privilege escalation in Windows Kernel-Mode Drivers allows an authenticated attacker who already holds low-privileged code execution on the host to elevate to a higher-integrity context, with total impact to confidentiality, integrity and availability. The CVSS vector (AV:L/AC:L/PR:L/UI:N/S:U) confirms the flaw is only reachable from a local session by a low-privileged principal, requires no user interaction, and needs no non-default feature to be enabled, so every unpatched build of Windows 10 (1607, 1809, 21H2, 22H2), Windows 11 23H2 and Windows Server 2012 through 2025 is exposed once an attacker has a foothold; there is no remote or network-reachable path. CISA SSVC records exploitation as none and no public exploit code has been identified at time of analysis; Microsoft has published a fix, making this a credible but not emergency-tier patching priority whose practical abuse case is post-access escalation on already-breached or multi-user hosts.
Authentication bypass in ZITADEL's Login V2 UI (versions 3.0.0 through 4.19.1) allows unauthenticated attackers to take over user accounts, including administrators with MFA completed, by forging unsigned session cookies and guessing time-ordered session IDs around a victim's sign-in to trigger token rotation. Exploitation requires the target to run the affected Login V2 UI and the attacker to hit a narrow timing window; no public exploit code is identified at time of analysis. While the CVE carries a CVSS 4.0 score of 9.2, the independent assessment rates it with high attack complexity (CVSS 3.1 AV:N/AC:H/PR:N/UI:N), reflecting the timing prerequisite.
Pre-authentication server-side request forgery in the SonicWall SMA1000 Appliance's Work Place interface lets a remote attacker who can reach the appliance abuse an unintended alternate access path that bypasses authentication and coerce the device into issuing requests on the attacker's behalf, reaching internal functionality that is not otherwise exposed externally (assessed CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:N - high confidentiality impact, low integrity impact, no availability impact). Affected builds are SMA1000 12.4.3-03526 (platform-hotfix) and older, and 12.5.0-02952 (platform-hotfix) and older; because the Work Place portal is normally internet-facing on these remote-access appliances, exposure is broad and this is a genuine top-priority vulnerability rather than a high-CVSS-but-low-risk artifact. No public exploit identified at time of analysis, and exploitation requires only default network reachability with no credentials and no user interaction, so the practical impact is bounded mainly by which internal services the appliance can reach through the forged requests.
Fork-originated Gitea Actions workflows can run on a base repository's registered runners without the required approval when a repository maintainer performs ordinary pull request triage, such as adding a label, because the approval decision was keyed to the user who triggered the event rather than the fork pull request author, allowing fork-controlled workflow code to execute in the base repository's CI context. Gitea versions up to and including 1.27.3 are affected when Actions is enabled and a matching runner is registered; the issue is fixed in Gitea 28.0.0. No public exploit code was identified at time of analysis and EPSS is low (0.15%, 3rd percentile), but exploitation is not autonomous: the assessed vector (CVSS:3.1/AV:N/AC:H/PR:L/UI:R/S:U/C:H/I:H/A:H) requires an attacker-submitted fork pull request plus a privileged maintainer action, and CISA SSVC rates the technical impact as total, so the outcome when the prerequisite chain is met is full compromise of confidentiality, integrity, and availability of the workflow execution environment.
Origin policy bypass in Google Chrome for Android prior to 155.0.8059.39 lets a remote attacker acting through a confused deputy have actions carried out under a different web origin's authority, so a victim who is socially engineered into opening a crafted HTML page can have that page reach resources or capabilities it should never hold for its own origin. Only Chrome on Android is affected, exploitation is not silent (it requires the user to open attacker-supplied content, CVSS PR:N/UI:R), and no public exploit code or active exploitation has been identified at time of analysis; EPSS is 0.22% (12th percentile), and Google rates the underlying Chromium bug as Low severity. Note the sharp signal conflict that must be reconciled before prioritization: the nominal vendor CVSS is 8.3 while EPSS, vendor severity, and our independent assessment (CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N) all point to materially lower real-world risk.
Confused-deputy bypass in Google Chrome's Contextual Tasks feature allows a remote attacker who already holds renderer code execution to cross a security boundary into a privileged page by loading a crafted HTML page, with high confidentiality and integrity impact (unauthenticated per the CVSS vector, but requiring user interaction). The flaw affects Chrome builds prior to 155.0.8059.39 and functions primarily as a chain link - the prerequisite renderer compromise from a separate bug is the real gate, which sharply narrows standalone risk despite the 8.1 score. No public exploit code has been identified at time of analysis and EPSS is very low at 0.18% (7th percentile), with CISA SSVC recording no observed exploitation; a vendor-released patch exists in 155.0.8059.39.
Sandbox escape and potential arbitrary code execution outside the browser sandbox in Google Chrome via a confused-deputy flaw in the Google Lens feature, affecting all desktop releases prior to 155.0.8059.39. Exploitation is not standalone: a remote attacker must first have compromised the Chrome renderer process (a separate renderer-stage bug), then convince the victim to load a crafted HTML page, after which the Lens component can be induced to perform a privileged operation that crosses the sandbox boundary on the renderer's behalf - consistent with the CVSS:3.1 vector AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H and high attack complexity. No public exploit code has been identified at time of analysis and there is no confirmed active exploitation (CISA KEV absent, SSVC exploitation: none), but the maximum technical impact rating means the issue is worth prioritizing within normal patch cycles; vendor-released patch: 155.0.8059.39.
Remote attackers can bypass system access restrictions and reach a privileged internal page in Google Chrome for Android versions prior to 155.0.8059.39 by way of a confused-deputy weakness in the SignIn component, which performs privileged navigation on behalf of an untrusted caller. Exploitation requires the victim to load attacker-controlled crafted HTML (for example, visiting a malicious or compromised page), needs no authentication or prior privileges, and is limited to the Android build on pre-patch versions. No public exploit code has been identified at time of analysis, EPSS is low at 0.22% (12th percentile), and the issue is not confirmed as actively exploited; a vendor patch (155.0.8059.39) is available, so this is a moderate-priority patch-and-move-on item rather than an emergency.