Sandbox escape in Google Chrome before 155.0.8059.39 via a heap-based buffer overflow (CWE-122) in ANGLE, Chrome's graphics translation layer: a remote attacker who has already compromised a sandboxed renderer process can use a crafted HTML page to execute arbitrary code outside the sandbox. It is a second-stage primitive rather than an initial-access bug - the assessed vector CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H (8.3) reflects high attack complexity, a required pre-existing renderer exploit chain, and active user interaction (UI:R), with the scope change S:C capturing the crossing of the sandbox boundary, even though Chromium rates the underlying bug Low severity. No public exploit code was identified at time of analysis and CISA's SSVC data records no observed exploitation, so the headline 8.3 should be weighed against the significant preconditions.
Heap-based buffer overflow in Google Chrome's font-handling code (CWE-122) lets an attacker who has already compromised the renderer process execute arbitrary code outside the sandbox when a victim renders a crafted HTML page containing malicious font data. All Chrome builds prior to 155.0.8059.39 on default configurations are affected; exploitation is unauthenticated (PR:N) but requires user interaction (UI:R) and a pre-existing renderer compromise, so it functions as a sandbox-escape chain component rather than a standalone drive-by, and is rated high complexity (AC:H) despite a CVSS base of 8.3 (S:C). No public exploit code has been identified at time of analysis and CISA's SSVC assessment records no exploitation, but the vendor has shipped a fix in 155.0.8059.39.
Sandbox escape in Google Chrome on macOS, affecting releases prior to 155.0.8059.39, allows an attacker who has already compromised the browser's renderer process to execute arbitrary code outside the sandbox when the victim views a crafted HTML page. The CVSS 3.1 vector (AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H, base 8.3) confirms a network-reachable, scope-changing escape that requires user interaction and carries high confidentiality, integrity and availability impact once triggered, but the AC:H rating and CISA SSVC 'Automatable: no' reflect that it cannot be used standalone - a prior renderer compromise is a hard prerequisite. Per the authoritative assessment this is a chained second-stage exploit limited to macOS; no public exploit code and no confirmed active exploitation were identified at time of analysis.
Type confusion in ANGLE, Chrome's OpenGL ES abstraction layer, lets a remote attacker who has already compromised the renderer process execute arbitrary code outside the browser sandbox in Google Chrome versions prior to 155.0.8059.39. The victim must load attacker-controlled graphics content (crafted HTML invoking WebGL/ANGLE paths), so user interaction is required, and the bug cannot be triggered by merely visiting a page without a separate renderer exploit first - it is a second-stage sandbox-escape primitive in a multi-bug chain rather than an initial-access vector. No public exploit identified at time of analysis; CISA's SSVC assessment rates exploitation as none and automatable as no, but technical impact as total, consistent with the CVSS scope change (S:C), high confidentiality and integrity impact, and the High Chromium security severity rating.
A sandbox escape in Google Chrome before 155.0.8059.39 lets an attacker who has already compromised the browser's renderer process trigger a race condition in Chrome's font handling and execute arbitrary code outside the sandbox, via a crafted HTML page the victim must load (CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H, score 8.3). The prerequisite is significant: a separate, prior renderer compromise is required, and AC:H reflects the non-deterministic nature of winning the race, so this is a chained post-exploitation primitive rather than a standalone drive-by. Unauthenticated from a network perspective, with UI:R meaning the target must open attacker-controlled content; no CISA KEV entry, no public exploit code identified at time of analysis, and CISA's SSVC assessment rates exploitation as none and automatable as no, though technical impact is total once the race is won. Vendor-released patch: 155.0.8059.39.
Use-after-free in Google Chrome's Storage component (all versions prior to 155.0.8059.39) allows a remote attacker to execute arbitrary code outside the sandbox, but only after the attacker has already compromised the renderer process - typically by chaining a separate renderer bug - and only when a victim loads a crafted HTML page (CVSS 8.3, AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H). Because a prior renderer compromise, user interaction, and non-trivial heap grooming are all required, this is a second-stage sandbox-escape primitive rather than standalone initial access. No public exploit code has been identified at time of analysis, CISA SSVC records exploitation as none, and Google has shipped a vendor patch in Chrome 155.0.8059.39.
Improper access control in the vertex result caching subsystem of IBM Langflow OSS 1.0.0 through 1.12.2 lets a remote authenticated attacker read sensitive cached data or inject malicious content into vertex results that other users then consume. The issue carries a CVSS 3.1 base score of 8.3 (High) under the vector AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:L, meaning it is network-reachable with low complexity and requires only a low-privileged account, while yielding high confidentiality and integrity impact and low availability impact. No public exploit code has been identified at time of analysis and the flaw does not appear in CISA KEV; IBM has published a patch advisory (node/7290694), so exposure is limited to unpatched deployments and to attackers who already hold valid credentials on the instance.
Arbitrary code execution outside the browser sandbox is possible in Google Chrome on macOS prior to 155.0.8059.39, where incorrect reference resolution in the WebAppInstalls component allows crafted network traffic to cross a process trust boundary. Exploitation is not standalone: the attacker must already have compromised the Chrome renderer process, which makes this a second-stage sandbox-escape primitive in a larger exploit chain rather than a directly reachable remote compromise, and the assessed vector (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:C/C:H/I:H/A:H) reflects that non-trivial precondition. A vendor-released patch (Chrome 155.0.8059.39) is available, and no public exploit code was identified at the time of analysis.
A race condition in Google Chrome's Fonts component prior to version 155.0.8059.39 lets a remote attacker escape the browser sandbox and execute arbitrary code at the browser's privilege level, but only after the attacker has already compromised a sandboxed renderer process through some separate, unrelated vulnerability. The exploit additionally requires the victim to load a crafted HTML page (UI:R) and the attacker to win a timing-dependent race (AC:H), so this is best understood as a sandbox-escape primitive rather than a standalone initial-access vector. There is no public exploit identified at time of analysis - CISA's SSVC assessment rates exploitation as 'none' and the flaw as not automatable - and a vendor-released patch (155.0.8059.39) is available, making this a high-priority defense-in-depth fix rather than a catastrophic unpatched bug.
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.
System access restriction bypass in Google Chrome versions prior to 155.0.8059.39 allows a remote attacker to break out of the browser's security boundary through the FullScreen component, but only after having already compromised the Chrome renderer process and only with successful social engineering to make the victim interact with a crafted HTML page. Because it is a chain component rather than a standalone bug, exploitation requires a pre-existing renderer foothold plus mandatory user interaction (UI:R) and carries high attack complexity (AC:H), so it is not exploitable against a fully patched default browser on its own. No KEV listing and no publicly available exploit code have been identified at time of analysis, and EPSS is low at 0.22% (12th percentile), consistent with the conditional nature of the flaw despite the 8.3 CVSS base score and a Chromium security severity rating of Medium.
Sandbox escape in Google Chrome versions prior to 155.0.8059.39: a race condition (CWE-362) in the browser's font handling can be triggered when a victim renders a crafted HTML page, potentially allowing a remote, unauthenticated attacker to execute arbitrary code outside the browser sandbox. Exploitation requires the target to load or view the attacker-controlled page (UI:R) and depends on winning a timing-sensitive race (AC:H), which typically demands repeated attempts and precise memory grooming, making this a credible high-priority browser flaw rather than a confirmed emergency. No publicly available exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis, and a vendor patch is available from Google.
Arbitrary code execution outside the browser sandbox is possible in Google Chrome on macOS prior to 155.0.8059.39, where a timing-dependent race condition (CWE-367) in the Core component can be won to break out of the renderer's sandbox. Exploitation is conditional rather than a single remote shot: the attacker must already have compromised the renderer process, must socially engineer the victim into opening and interacting with a crafted HTML page (UI:R), and must win a non-deterministic race that the AC:H rating and CISA's SSVC assessment (Exploitation: none, Automatable: no) indicate is unreliable and not readily scriptable. No public exploit code has been identified at time of analysis, and Google has shipped a fixed build, 155.0.8059.39; because the mandatory pre-existing renderer compromise is a high bar, this is best treated as a conditional priority to be handled through normal patching cadence rather than an emergency.
Cross-site scripting in @quasar/app-vite (Quasar Framework) SSR and SSG rendering paths prior to 3.3.0 allows HTML attribute and markup injection when an application derives or overrides ssrContext.nonce from attacker-controllable data instead of relying on Quasar's default cryptographically generated base64/base64url nonce. The flaw is exploitable remotely and unauthenticated in the sense that no privileges are required on the server side (CVSS PR:N), but it is inert under default configuration: a quote character must survive into the nonce value, and the poisoned server-rendered page must then be loaded by a victim, so exploitability is contingent on the application's own nonce handling (independent assessment rates it AV:N/AC:H and requires user interaction). There is no confirmed active exploitation (not in CISA KEV) and no public exploit code identified at time of analysis; the issue is fixed in version 3.3.0.
Apache Commons BCEL before 6.13.0 caches attacker-supplied class files under the symbolic name declared inside the class file itself rather than validating that name against the name it was requested under, so a class loaded via AbstractClassPathRepository or ClassLoaderRepository can poison a repository entry keyed to a different, colliding name. Applications that load class files from an attacker-influenceable classpath or input stream and later rely on the repository's name-keyed lookups or verification results can be induced to operate on the wrong class, producing an integrity-only impact (no confidentiality or availability effect). The independent assessment rates this CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:H/A:N - network-reachable and unauthenticated, but requiring a specific name-collision and subsequent lookup sequence (AT:P), so it is not a fully automatable drive-in exploit; no public exploit code and no confirmed active exploitation were identified at time of analysis, and the vendor's elevated 8.2 CVSS 4.0 score reflects the integrity impact alone.
Cross-origin data disclosure in Google Chrome (all versions prior to 155.0.8059.39) stems from a missing authorization check in the browser's FileSystem component, allowing a crafted HTML page to breach the web origin boundary and read data it should not reach. Exploitation is strictly a second-stage condition: an attacker must first have already compromised Chrome's renderer process through a separate flaw, and the victim must load attacker-controlled content (UI:R), so a default, fully patched Chrome with an intact renderer sandbox is not directly exploitable - a limitation that makes real-world risk moderate despite the 8.2 base score, and is reflected in the low EPSS of 0.22% (12th percentile). Google has shipped a fix in Chrome 155.0.8059.39, there is no confirmed active exploitation (CISA KEV) and no public exploit identified at time of analysis.
Site isolation bypass in Google Chrome for Android (versions prior to 155.0.8059.39) lets an attacker who has already compromised the renderer process cross the per-site process boundary and reach data that should remain isolated to other origins, triggered when the victim loads a crafted HTML page. It is a second-stage flaw rather than an initial-access vector: exploitation is gated on a separate, pre-existing renderer compromise plus required user interaction (CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:L/A:N), which is why the vendor assigned a Chromium severity of Low even though the base score is 8.2. No public exploit identified at time of analysis, no CISA KEV listing, and EPSS is only 0.21% (11th percentile); a vendor patch is available, so the residual risk is concentrated in unpatched Android clients that are simultaneously exposed to a first-stage renderer exploit.
Privilege escalation in Dell System Update versions prior to 2.3.0.0 lets a low-privileged local attacker abuse an incorrectly secured critical resource to gain higher privileges (CVSS 8.2, CWE-732). Exploitation requires local access, a valid low-privileged account, and user interaction-most plausibly an administrator running the updater after malicious content has been staged in a writable location-so remote or unauthenticated attack is not possible. No public exploit code or confirmed active exploitation (CISA KEV) was identified at time of analysis.
Elevation of privilege in Dell System Update (DSU) versions prior to 2.3.0.0 allows an authenticated, low-privileged local attacker to reach a privileged code path and operate at higher integrity than intended, stemming from an improper access control weakness (CWE-284, tagged by the reporter as an authentication bypass). Exploitation is conditional rather than drive-by: the attacker must already hold a local account on the endpoint (AV:L, PR:L) and a separate user interaction step (UI:R) is required to trigger the vulnerable path, so this is a post-foothold escalation scenario with no remote or network-reachable vector. CVSS 3.1 scores it 8.2 with a changed scope and high confidentiality, integrity and availability impact, but no public exploit code has been identified at time of analysis and the issue is not listed in CISA KEV.
Ciphertext tampering in the Cloak Elixir encryption library (danielberkompas/cloak, all versions from 0.1.0-pre onward) lets anyone with write access to the encrypted store silently flip bits in stored ciphertext so that decryption returns an attacker-chosen plaintext of the same length. The affected code paths, Cloak.Ciphers.AES.CTR and Cloak.Ciphers.Deprecated.AES.CTR, encrypt with AES-256 in CTR mode and store the key tag, IV, and ciphertext with no MAC, and decrypt/2 validates only the key tag and a minimum length - so tampered data is returned as valid with no error. Exploitation requires write access to the ciphertext (the description cites SQL injection into the Ecto-backed database or a compromised replica) plus knowledge or a correct guess of the existing plaintext at the tampered offset, and the replacement value must be exactly the same length as the original; impact is integrity-only (VC:N/VA:N), with no public exploit identified at time of analysis. Note the vendor CVSS 4.0 score of 8.2 is driven entirely by VI:H, while an independent assessment applying CVSS 3.1 (AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:H/A:N) rates the realistic risk considerably lower because an attacker must already control writes to the encrypted store and have plaintext knowledge.
Remote code execution in Payload CMS before 3.90.0 (and canary builds before 4.0.0-canary.34) lets a remote, unauthenticated attacker execute arbitrary code by sending a crafted request to the public first-register operation. The flaw only applies to instances that have local authentication enabled and have not yet had their first admin user created, i.e. a freshly deployed, re-provisioned, or ephemeral instance still inside its initial setup window; once an admin exists the attack surface disappears. No public exploit identified at time of analysis, and no evidence of active exploitation (not in CISA KEV), though the CVSS 8.1 score with full confidentiality, integrity and availability impact makes this a high-priority fix for anyone running uninitialized deployments.
Authenticated arbitrary file deletion in Payload CMS before version 3.90.0 (and canary releases before 4.0.0-canary.34) lets a user who can update or delete uploads stored on the local filesystem cause the file-cleanup routine to remove files outside the configured upload directory, resulting in data loss or service disruption. The flaw is a path traversal in the upload deletion logic that requires a valid account holding update/delete permission on an upload-enabled collection (CVSS 3.1: AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:H, score 8.1), so it is not exploitable by anonymous visitors. There is no public exploit identified at time of analysis and the issue is not reported as actively exploited; deployments using cloud/object storage adapters or restricting upload management to trusted operators are largely unaffected. Vendor-released patches are available in 3.90.0 and 4.0.0-canary.34.
Authenticated users of Backstage Scaffolder can bypass administrator-configured action restrictions in @backstage/plugin-scaffolder-backend by supplying a restricted input value with different letter casing or as a different JSON type (for example a number instead of a string), causing permission checks to fail to match and letting the action execute with the backend integration's credentials. This can grant unauthorized access to repositories and related source-control resources, bounded by the privileges of the integration tokens attached to the affected actions. The issue affects versions before 3.3.1, 3.4.1, 4.0.3, and 4.1.0, and only matters in deployments that actually use input-value-based permission rules; no public exploit code has been identified at time of analysis.
Authenticated remote code execution in IBM Langflow OSS 1.0.0 through 1.12.2 lets any holder of valid credentials - even a low-privileged one - submit or edit a flow that reaches the server-side code-generation and execution path, resulting in arbitrary OS command execution on the Langflow backend. IBM scores it 8.1 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N) and classifies the root cause as CWE-94 code injection; no public exploit code has been identified at time of analysis and there is no CISA KEV listing, so active exploitation is not confirmed. The decisive risk factor is deployment posture: internet-exposed or multi-tenant instances running with authentication disabled, default or shared credentials, or open self-registration collapse the PR:L barrier and behave like unauthenticated RCE targets.
Unauthenticated privilege escalation in the GiveWP WordPress donation plugin (versions up to and including 4.17.0) lets a remote attacker gain elevated privileges on a vulnerable site without any credentials or user interaction, consistent with the assessed vector CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H. The key mitigating nuance is the High attack complexity: success depends on preconditions the attacker cannot fully control, so this is a serious 8.1 High-severity issue rather than a trivially scriptable one. No public exploit code has been identified at time of analysis and the CVE is not listed in CISA KEV, but because it is unauthenticated and grants high confidentiality, integrity, and availability impact, sites running GiveWP 4.17.0 or earlier should treat it as urgent; sites that do not run the plugin or have already moved past 4.17.0 are not affected.
Improper authentication in the Backstage @backstage/plugin-auth-backend-module-oidc-provider package prior to 0.4.20 lets an authenticated user of a configured OIDC identity provider impersonate another catalog identity. Exploitation requires a deployment that resolves sign-in identities by matching the OIDC email claim AND an identity provider that issues email addresses without asserting email_verified=true, in which case the attacker's unverified, self-controlled email can match a victim's catalog identity and inherit that identity's access and permissions (CVSS 3.1 base 8.1, PR:L, no availability impact). No public exploit code has been identified at time of analysis, and the flaw is fixed by omitting the email from the default OIDC profile transform unless the provider explicitly marks it verified.
Insufficient audience validation in Backstage's Cloudflare Access auth provider package (@backstage/plugin-auth-backend-module-cloudflare-access-provider 0.1.0 through 0.4.x) lets a user who already holds a valid Cloudflare Access JWT issued for a different Access application inside the same Cloudflare Zero Trust team authenticate to Backstage, because the provider verifies the token signature and team issuer but never checks the token's audience (AUD) claim. Exploitation is conditional: the forged-context token must actually reach the Backstage auth endpoint on a path where Cloudflare Access has not already enforced the Backstage application's own AUD tag upstream, and the attacker must be an authenticated member of the shared Zero Trust team, which is why the assessed vector carries AC:H and PR:L and the resulting rating is a moderate, conditional authentication bypass rather than a top-tier priority. No public exploit code has been identified at time of analysis, and the issue is fixed in plugin version 0.5.0 (shipped with Backstage v1.55.0), where the 'audience' configuration value becomes mandatory.
Access-control bypass in SSSD lets an already-authenticated user from a trusted identity domain reach hosts and services that Host-Based Access Control (HBAC) rules are supposed to protect. Because SSSD strips the domain qualifier and compares only short usernames during HBAC evaluation, a trusted-domain user whose short name collides with an authorized local account is silently granted that local account's access. Publicly available exploit code exists, but exploitation is not automatable and EPSS is low at 0.28% (18th percentile); exposure is limited to trust-enabled FreeIPA/IdM-style deployments that delegate host access decisions to SSSD.
Unauthenticated PHP object injection in the Haaken WordPress theme affects all versions up to and including 1.5, allowing remote attackers to supply crafted serialized data to an unserialize() sink without any prior authentication (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H, base 8.1). The conditional nature of the risk is captured by the AC:H rating: full impact (remote code execution or data compromise) generally depends on a usable POP gadget chain being present in the theme, WordPress core, or another installed plugin, so exposure is limited to sites actually running this theme at an affected version. No public exploit code or confirmed active exploitation (CISA KEV) was identified at time of analysis, and several vectors beyond the single Patchstack advisory reference (EUVD-2026-93080) remain unpublished.
Unauthenticated privilege escalation in the SMS Alert Order Notifications WordPress plugin through 4.0.0 allows network attackers to obtain elevated privileges on affected sites, with the CVSS 3.1 vector AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H (base 8.1 High) indicating no authentication or user interaction but high attack complexity. The underlying issue is CWE-266 incorrect privilege assignment, and successful exploitation depends on a condition outside the attacker's control such as a race condition, predictable value, or specific plugin/site state; the exact trigger was not disclosed in the provided data. No public exploit identified at time of analysis.
Site isolation bypass in Google Chrome on Android (prior to 155.0.8059.39) allows an attacker who has already compromised the renderer process to reach the browser's Passwords component via a crafted HTML page and a missing authorization check, exposing stored credential data. The bug is a chained, second-stage issue: it cannot be triggered by an unprivileged remote attacker without pre-existing renderer code execution, and it requires user interaction to load the crafted page, which is why the assessed vector (AV:N/AC:H/PR:L/UI:R/C:H/I:N/A:N) makes realistic risk LOW despite the 8.1 base score. No public exploit identified at time of analysis, EPSS is only 0.21% (11th percentile), CISA SSVC records Exploitation: none, and Google rated the Chromium security severity Medium, so this is a routine patch-cycle fix rather than an emergency.
A use-after-free write in the Linux kernel's lan743x Ethernet driver (Microchip LAN7430/LAN7431) lets a packet that spans more than one RX buffer trigger a kernel memory corruption through a stale skb pointer in the checksum-success path of lan743x_rx_process_buffer(), with the flawed write landing on freed fragment skb metadata. According to the vendor description this multi-buffer condition only arises transiently after an operator performs a live MTU increase, because existing ring entries retain their old, smaller buffer size until they are replenished; the vulnerable driver must be present and the host must be using that specific Microchip hardware. Vendor patches are available in the stable trees (6.1.189, 6.6.158, 6.12.112, 6.18.54, 7.2.8 and 7.3-rc4), EPSS is low at 0.18% (7th percentile), and no public exploit code or CISA KEV activity was identified at time of analysis; our independent assessment rates real-world priority LOW despite the vendor-supplied 8.1 CVSS because exploitation is gated on hardware, a concurrent operator-initiated MTU change, and a short-lived window.
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.
Site isolation bypass in Google Chrome versions earlier than 155.0.8059.39 lets a remote attacker who has already compromised a renderer process escape Chrome's process boundary and reach data or privileged accessibility functionality belonging to other origins by way of a crafted HTML page the victim loads. The flaw is an incorrect-authorization bug (CWE-863) in the browser's Accessibility component, making it a chained, second-stage primitive rather than a standalone remote compromise, which is why the assessed vector uses AC:H (AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:N) despite an underlying base score of 8.1. Google has released a vendor patch in 155.0.8059.39; no public exploit code and no confirmed active exploitation were identified at time of analysis (EPSS 0.18%, CISA SSVC exploitation: none, automatable: no), so the risk is real but is not currently a frontline emergency.
Kernel memory corruption in the Linux CIFS/SMB client (cifs.ko) allows a malicious or compromised SMB server to have a channel and its primary server structure freed while the client still dereferences them, producing a use-after-free that crashes the kernel and can expose or corrupt freed memory. Exposure is limited to clients that mount a share from an attacker-controlled server with SMB multichannel in use: the vector (AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:H) indicates no client-side credentials are required, but the victim must initiate the mount and the bug depends on a race/teardown timing window, so the net real-world risk is lower than the 8.1 base score suggests. No public exploit code is identified at time of analysis, EPSS is 0.18% (6th percentile), and vendor-released patches exist for the affected stable branches.
Broken access control in Gitea up to and including 28.0.0 allows an authenticated repository collaborator who holds write access to the Code unit (but not Releases write access) to permanently delete published releases and their attachments by sending a single POST to /{owner}/{repo}/tags/delete against the release's tag. The flaw is rated CVSS 8.1 (AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:H per the assessed vector), reflecting high integrity and availability impact with no confidentiality loss, and exploitation requires the attacker to already be a trusted collaborator on the target repository rather than an anonymous outsider. There is no public exploit code identified at time of analysis, CISA KEV does not list it, SSVC rates exploitation as none and automatable as no, and EPSS is low at 0.17% (6th percentile).
Site isolation bypass in Google Chrome affects versions prior to 155.0.8059.39, where a missing authorization check (CWE-862) in the browser's Workers implementation lets an attacker who has already compromised a renderer process cross the site-isolation boundary using a crafted PDF file. The attack is remote (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N, score 8.1), but exploitation is gated on a pre-existing renderer foothold obtained through a separate vulnerability, so this is a chained sandbox/site-isolation escape rather than a standalone remote compromise. There is no public exploit identified at time of analysis and no CISA KEV listing; EPSS is low at 0.17% (5th percentile), and CISA SSVC rates exploitation as 'none' with automation as 'no', while acknowledging total technical impact if the chain is completed. A vendor patch is available in Chrome 155.0.8059.39.
Blind SQL injection in LogicalDOC Enterprise up to and including 9.1.1 lets an authenticated user inject arbitrary SQL through the workflow template name parameter processed by the WorkflowsDataServlet, potentially exposing or tampering with the backend database that stores documents and metadata. Exploitation requires a valid account that can create or submit a workflow template (PR:L; no administrative role, no victim interaction), so realistic exposure is limited to insiders or attackers who already hold credentials against a niche document-management deployment. EPSS is low at 0.17% (5th percentile), the product is not in CISA KEV, and no public exploit has been identified at time of analysis.
A teardown race in the Linux kernel's iSER target (ib_isert) frees the isert_conn and iscsit_conn structures while a deferred control PDU completion is still queued on isert_comp_wq, so isert_put_cmd() later dereferences freed memory - a slab use-after-free (KASAN splat) rated CVSS 8.1 (AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H) because it can in principle lead to kernel code execution, data disclosure, or a crash. Exposure is limited to hosts that run the LIO iSER target with the ib_isert module loaded and export an iSER/RDMA transport over an InfiniBand or RoCE fabric; the CVSS vector shows no authentication prerequisite (PR:N), but the attacker must win a narrow race between connection teardown and the workqueue handler, a window so tight that the reporter needed an injected msleep(200) to reproduce it reliably. No public exploit code and no CISA KEV entry are associated with this issue, EPSS is only 0.16% (5th percentile), and vendor-released stable patches are available.
Gitea through 1.27.3 fails to revoke per-unit repository access when an organization team is demoted through the REST API, so existing team members silently keep their previous unit permissions - including write access to the team's repositories - after an owner lowers the team level (for example admin to read). The flaw is CWE-272 Incorrect Privilege Assignment: a PATCH that supplies only the permission field updates the team's access mode but never rebuilds its per-unit grants nor applies the new level as a cap, and exploitation requires an authenticated organization owner to make that API call plus a team that already holds per-unit repository permissions, with the CVSS vector (PR:L) confirming an authenticated beneficiary; the web form path is not affected. Vendor scoring is 8.1 with high confidentiality and integrity impact, while EPSS is only 0.16% (5th percentile) and no public exploit code has been identified at time of analysis; upgrading to Gitea 28.0.0 restores the intended cap behavior.
Unauthenticated attackers who can reach Gitea's web installer on an instance whose INSTALL_LOCK has been reset to false (or is otherwise unset) can obtain a fully authenticated session for any existing account - including an administrator - simply by submitting the install form with a username that matches that account, with no password verification whatsoever. Full administrative takeover follows, allowing the attacker to change the account's password and control the instance. The bypass only exists while the installer is network-reachable against a database that already contains users; the default hardened post-install state (INSTALL_LOCK=true) is not affected. Independent assessment rates this CVSS 8.1 (AV:N/AC:H/PR:N/UI:N, C:H/I:H/A:H); EPSS is low at 0.16% (4th percentile) and no public exploit code has been identified at time of analysis.
Push mirror synchronization in Gitea 1.27.3 and earlier reuses a remote address that was validated against the [migrations] allow/block lists only once, at mirror-creation time, so a hostname that later resolves to a blocked or internal address is still handed to git push - a time-of-check/time-of-use flaw (CWE-367). An authenticated user who administers a repository (which includes any repository they create themselves) can therefore aim a push mirror at an internal Git service and force-push the repository's contents to it, assuming attacker influence over name resolution and a network-reachable internal endpoint. The assessed severity is high (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:N) because scope changes and both confidentiality and integrity of internal Git services are affected, but no public exploit code has been identified at time of analysis and EPSS is low (0.12%, 2nd percentile), with SSVC recording no observed exploitation and a non-automatable attack path - a real but bounded risk rather than a mass-exploitation emergency.
Remote code execution (RCE) in WPShopmart Tabs Responsive plugin for WordPress (versions <= 2.5) allows an authenticated attacker with Editor-level privileges to execute arbitrary code on the server. This high-severity vulnerability (CVSS 8.0, CWE-94) requires high-complexity conditions (AC:H) and no user interaction (UI:N). No public exploit code has been identified at time of analysis; risk is concentrated on multi-user sites where untrusted users have elevated editing roles.
Unauthenticated SQL injection in the eLoanApp Platform's AJAX user-verification endpoint allows remote attackers to inject boolean- and time-based payloads into the POST parameter 'logina', enabling database engine discovery and query delays. The affected product is RDL Technologies eLoanApp Platform as represented by CPE cpe:2.3:a:rdl_technologies:eloanapp_platform:*:*:*:*:*:*:*:*, and the independent assessment scores this as CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:L, meaning network-reachable, unauthenticated exploitation with low confidentiality, integrity, and availability impact rather than full compromise. No public exploit identified at time of analysis, and the practical risk is moderate because blind extraction is slow and noisy, making large-scale data theft time-consuming and more detectable than error-based injection.
Configuration-controlled code execution in the hydra-optuna-sweeper package (Hydra 1.2.0 up to but not including 1.3.0, plus 1.4.0 development builds before 1.4.0.dev10) allows an attacker who controls Optuna sweep configuration or command-line overrides to name an arbitrary importable Python callable in hydra.sweeper.custom_search_space that Hydra resolves and then invokes inside the controller process at the application's privilege level, including bypassing the trusted execution whitelist introduced on affected 1.4.0 development releases. The assessed CVSS 3.1 vector is AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H (7.8 High): delivery is local and unauthenticated, but a victim must actually launch the sweep, after which confidentiality, integrity, and availability impact is total for the affected process. Exploitation is only meaningful where sweep configuration or overrides cross a trust boundary; no public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis.
Use-after-free in the Linux kernel netfilter flowtable datapath allows a local, authenticated low-privileged user to potentially execute code, corrupt memory, or disclose information by winning a race condition against the RCU grace period. Affected are unpatched kernels in the 5.10.x, 5.15.x, 6.1.x, 6.6.x, 6.12.x, 7.2.x, 7.3-rc, and 6.18.x ranges before their respective fixes; systems without a configured netfilter flowtable offload are not exposed. No public exploit code has been identified at time of analysis, and the EPSS score is 0.22% (11th percentile), indicating low predicted exploitation activity; vendor patches are available.
Local use-after-free in the Linux kernel's virt_wifi virtual wireless driver (CVE-2026-98311) lets a low-privileged local attacker with network-admin capability corrupt kernel memory, causing a crash and potentially enabling code execution in kernel context. virt_wifi_newlink() propagates the lower device's operstate before the new netdevice is registered, so a dormant lower device queues the still-uninitialized netdev on the linkwatch event list; if registration later fails - attacker-controllable by supplying an interface name such as 'bad/name' - free_netdev() releases the object and the subsequent linkwatch_fire_event() dereferences freed memory. Exposure is narrow: exploitation is strictly local, requires CAP_NET_ADMIN (attainable via user namespaces on some distributions) plus the rarely deployed virt_wifi module, no public exploit code has been identified, CISA KEV does not list this issue, and EPSS is only 0.22% (11th percentile).
Local privilege escalation and code execution in Dell System Update (DSU) versions prior to 2.3.0.0 allows an authenticated, low-privileged local user to supply a crafted pathname that escapes the directory DSU intends to restrict, causing the elevated DSU process to read, write, or execute content from an attacker-controlled location. Exploitation yields complete loss of confidentiality, integrity, and availability on the affected host (CVSS 7.8, AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H), but it is not remotely reachable and requires the attacker to already hold a local low-privileged account or code-execution foothold; no public exploit code has been identified at time of analysis, EPSS is low at 0.22% (11th percentile), and CISA's SSVC assessment rates exploitation as 'none' with automatable 'no'. Vendor-released patch: 2.3.0.0.
Use-after-free in the Linux kernel's IPsec/xfrm receive path (xfrm_input) can be triggered by a local attacker with low privileges on hosts that run IPsec with VTI tunnels or XFRM interfaces configured. The receive callbacks for VTI and XFRM interfaces - as well as the XFRM_MAX_DEPTH error path - call secpath_reset(), which can drop the last reference to an xfrm_state that xfrm_input() subsequently dereferences when selecting transport_finish() or reporting the drop-callback protocol; if that state is concurrently deleted, the freed object is accessed, yielding memory corruption with high confidentiality and integrity impact. Successful exploitation requires winning a race between secpath_reset() and concurrent xfrm_state teardown, so the assessed vector is CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H (note the high attack-complexity, which differs from the NVD-published AC:L value); EPSS is low at 0.21% (10th percentile), there is no KEV listing, and no public exploit identified at time of analysis.
Deserialization of untrusted data in NVIDIA Model-Optimizer (CVSS 3.1 base score 7.8) allows an attacker who supplies a crafted model artifact to achieve code execution, data tampering, denial of service, and information disclosure on the machine that loads it. The assessed vector (AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H) shows the severity comes from full loss of confidentiality and integrity once the payload runs, but the attacker must first get the victim to open or deserialize the malicious artifact in a local processing environment, so there is no remote, zero-interaction path. No public exploit code has been identified at time of analysis and the CVE is not listed in CISA KEV, so this is a credible but priority-setting issue for ML/AI pipelines that ingest third-party models rather than an emergency remote-exploitation case.