CordysCRM 1.7.3 and earlier allows an authenticated user holding the PROCESS_SETTING_ADD permission to store an attacker-chosen webhook URL in an approval-flow node and then cause the server to issue outbound HTTP requests to that URL when an approval action is approved, turning the application into a proxy for reaching internal-only services. The assessment rates this CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:N/A:N - high confidentiality impact against the attacked component with no integrity or availability impact - and the practical payoff is cloud metadata access (for example 169.254.169.254) and internal network reconnaissance, bounded by the server's network reach and by the multi-step trigger (configure via POST /approval-flow/add, fire via POST /approval-action/approve). The flaw is fixed in 1.7.4 per vendor advisory GHSA-fg6q-pfj7-fghw, PR #2966 and commit e0ae23e; no public exploit code and no confirmed active exploitation were identified at time of analysis, and exploitation requires a low-privileged authenticated account plus a second in-application approval step.
FluidSynth builds compiled with the native DLS parser enabled can be driven into a local denial of service by a crafted DLS SoundFont file: an integer overflow in the parser's articulation-chunk size check lets the loader run roughly a billion out-of-bounds 12-byte iterations, causing a crash or hang. The flaw affects versions 2.5.0 through 2.5.5 and is fixed in 2.5.6; exploitation requires the target to be built with the non-default CMake option enable-native-dls set to ON and to load the attacker-supplied file (unauthenticated per the CVSS PR:N vector, but the attack vector is local and the malicious file must reach the application). No public exploit code or confirmed active exploitation was identified at time of analysis, and impact is confined to availability - there is no code execution or data disclosure.
A 32-bit integer overflow in FluidSynth's native DLS parser lets a crafted DLS sound-bank file bypass the ptbl chunk-size check, triggering an approximately 4 GB pool-cue allocation and billions of out-of-bounds reads that crash the process, producing a denial of service against the affected 2.5.0 through 2.5.5 releases. Exploitation requires the library to be built with enable-native-dls=ON and an attacker-supplied DLS file to be opened by the application; the CVSS vector indicates no privileges are required (PR:N), and no public exploit code was identified at time of analysis. The vendor has fixed the flaw in FluidSynth 2.5.6.
Denial of service in WinFsp through 2.2.26215 allows an authenticated, unprivileged local user to crash the Windows kernel by issuing FSP_IOCTL_TRANSACT requests to the WinFsp control device, triggering a null pointer dereference in the driver's Fast I/O device control handler. The vulnerability requires local access (PR:L) and a system with WinFsp installed; it results only in a system crash with no data exposure or code execution. No public exploit code has been identified at time of analysis, and the flaw is not known to be actively exploited.
Privileged-mode password recovery affects PLANET IGS-5225-8P2T4S industrial managed switches running V1 firmware before 1.2412b260707 or V2 firmware before 2.2412b260519, because these devices store administrative credentials using the weak MD5 hashing algorithm. An attacker who obtains the device configuration file - typically via an admin config export, an exposed backup, or physical/local device access - can crack the stored hash offline and recover the privileged-mode password. Exploitation is not remote and requires an authenticated, high-privileged foothold or local access to the config file (CVSS 3.1 AV:L/AC:L/PR:H/UI:N, confidentiality-only impact), and no public exploit has been identified at time of analysis; this is a genuine but low-urgency information-disclosure weakness rather than a critical remotely-reachable flaw.
Missing authorization on an appointment-retrieval action in the Bookit - Booking & Appointment Calendar WordPress plugin before 2.6.0.5 lets an authenticated user holding a low-privilege Bookit-specific role pull other customers' appointment records, including names, email addresses, phone numbers and private booking comments. The underlying issue is a CWE-200 information-exposure pattern: the action returns booking PII without verifying that the caller owns the appointment or holds the capability that normally gates such data. Publicly available exploit code exists and WPScan published the finding, but there is no indication of active exploitation, impact is confined to confidentiality with no write or availability effect, and a vendor fix is available in 2.6.0.5 - making this a low-to-moderate priority disclosure issue rather than a high-severity threat.
Deletion of arbitrary appointment records is possible in the WordPress plugin Bookit - Booking & Appointment Calendar in all versions before 2.6.0.5, because one of its appointment-deletion functions omits a capability check (CWE-862 Missing Authorization). Exploitation is authenticated: an attacker must already hold the plugin's custom low-privileged Staff role - a non-default WordPress role that an administrator has to provision - and can then invoke the deletion routine to remove bookings they should not be able to touch. Publicly available exploit code exists (referenced through the WPScan advisory), but this is not confirmed as actively exploited (no CISA KEV listing), and given the Staff-role prerequisite and the integrity-only, appointment-record-limited impact, it is a low-priority insider or multi-staff-deployment risk rather than an internet-wide unauthenticated threat.
MasterStudy LMS WordPress plugin before 3.7.50 fails to verify that an Instructor-role user owns a course before returning its enrolled-student data, allowing disclosure of student names and email addresses for other instructors' courses. Publicly available exploit code exists, and vendor has released patched version 3.7.50. The independent assessment rates this as genuinely low severity (CVSS 2.7) because exploitation requires an authenticated Instructor account and only exposes limited personal data with no write or privilege-escalation impact.
King Addons for Elementor versions before 51.1.81 fail to enforce capability, post-status, or password checks when rendering the content of a requested post, allowing any authenticated user with Contributor-level access or higher to read the contents of private, draft, pending, and password-protected posts they are not authorized to view. The flaw is an authenticated information-disclosure issue (CWE-200) that requires only an active contributor-level account on a site running the affected plugin plus the choice of a target post ID; it grants no write, modification, or destructive capability. Publicly available exploit code exists, but there is no indication of confirmed active exploitation and the vendor has shipped a fix in 51.1.81; our independent assessment rates this a genuinely low-priority, though real and patch-worthy, authorization flaw (assessed CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N).
Improper exception handling in pg_partman's run_maintenance() function allows an authenticated database role holding partman_user access to permanently stall automated partition maintenance across an entire database. Because the exception handler wraps the whole row-processing loop over part_config rather than each individual partition set, a single crafted row that reliably raises an error - and that is given a low maintenance_order value so it is processed first - immediately aborts the loop and skips every legitimate partition set scheduled after it. Each subsequent maintenance tick (background worker or cron) fails the same way, so child partitions stop being created and retention stops dropping old ones, producing a database-wide availability impact; no public exploit code has been identified at time of analysis and there is no KEV listing, though EPSS and exploitation data were not available in this feed. Fixes are available in pg_partman 5.5.0.
Unvalidated OIDC redirect handling in the Kyoo self-hosted media server before 5.1.0 lets an unauthenticated attacker capture a victim's one-use login token and exchange it for that victim's Kyoo session. The attacker starts the OIDC login flow with an attacker-controlled redirectUrl; because the pre-fix code stored that URL alongside the opaque login state and appended the token, provider, and error values to it without validating the destination or binding the token to the browser session that began the flow, a social-engineered victim who completes authentication at their identity provider has the token delivered to the attacker's host, where it can be replayed through /auth/oidc/callback/{provider}. CVSS 3.1 base score is 6.5 (AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:N/A:N) - network reachable and requiring no privileges or authentication, but gated on victim user interaction and on the instance having at least one OIDC provider configured and enabled; no public exploit identified at time of analysis.
HTTP request smuggling in Netty's HTTP/1.1 decoder (HttpObjectDecoder and the HttpRequestDecoder built on it) lets remote, unauthenticated attackers send a chunk-size line containing an embedded bare carriage return - for example `0\rX` - because strict chunk-line validation is skipped whenever the size line carries no `;` chunk extension, causing a request-boundary desync when Netty shares a connection with a second HTTP intermediary that frames the message differently. The flaw reaches every application and Red Hat product that embeds Netty, including AMQ Broker 7, AMQ Clients, Quarkus, Camel 4 for Quarkus 3, Camel for Spring Boot 4, Keycloak / Red Hat Single Sign-On, JBoss EAP 7 and 8, Fuse 7, Data Grid 8, Apicurio Registry 3 and Debezium 3, while independent assessment keeps it at CVSS 6.5 with AC:H because a standalone Netty service with no disagreeing proxy in the chain yields little practical smuggling impact. The public report includes a reproducible PoC test harness, but there is no CISA KEV listing and no evidence of active exploitation.
IBM QRadar deployments running 7.5.0 through 7.5.0 UP15 Interim Fix 006 expose sensitive backup-file contents to any holder of a valid QRadar account, because those backup files are stored with incorrect permissions. Exploitation requires authentication (CVSS vector PR:L) but no administrative rights and no user interaction, and an attacker who satisfies that bar can read configuration, credential, or event data captured in the affected backups, yielding a high confidentiality impact (C:H) with no integrity or availability effect. This is a moderate-priority information-disclosure issue rather than a high-urgency emergency: no public exploit code has been identified at time of analysis, the flaw is not in CISA KEV, and the authentication prerequisite plus the need for an unpatched interim-fix level are the principal limiting factors.
Adjacent-network attackers can drive the Name Service component of IBM WebSphere Application Server 8.5 and 9.0 into a denial-of-service condition by supplying maliciously crafted serialized data to the CORBA/IIOP naming endpoint, which the JVM then deserializes. The flaw is unauthenticated and requires no user interaction (CVSS:3.1/AV:A/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H, 6.5), so the practical barrier is network position rather than credentials: the attacker must be able to reach the Name Service on the same or an adjacent network segment, and default configurations leave that service listening on the network. Impact is confined to availability - no loss of confidentiality or integrity is assessed - and no public exploit code or confirmed active exploitation has been identified at time of analysis; IBM has published a vendor fix.
HTTP request smuggling in Netty's netty-codec-http component lets unauthenticated remote attackers bypass front-end security controls by sending a crafted HTTP/1.1 chunk-size token containing whitespace after the hexadecimal digits, which Netty parses differently from other HTTP intermediaries. The flaw carries a CVSS 3.1 base score of 6.5 (AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:L), meaning no authentication or user interaction is required, but integrity and availability impact is limited; meaningful exploitation generally depends on a multi-tier topology where a proxy, load balancer, or backend parses the chunk-size token inconsistently with Netty. Red Hat ships the affected library across a broad portfolio - including AMQ Broker 7, AMQ Clients, Quarkus, Keycloak and Red Hat Single Sign-On 7, JBoss EAP 7/8, Fuse 7, Data Grid 8, Camel for Spring Boot 4, Camel 4 for Quarkus 3, Apicurio Registry 3 and Debezium 3 - and no public exploit code has been identified at time of analysis, with no confirmed active exploitation (CISA KEV).
Server-side request forgery in OpenCVE versions prior to 3.0.0 lets an authenticated user who holds permission to configure notification channels force the OpenCVE server to issue HTTP or HTTPS requests to arbitrary attacker-chosen destinations, including localhost interfaces, RFC1918 internal ranges, link-local addresses and the cloud metadata endpoint 169.254.169.254, with the response content retrievable by the attacker. The prerequisite is the notification-configuration privilege (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N, CWE-918), so anonymous actors cannot trigger it, and the practical impact is highest in cloud-hosted deployments where an instance metadata service is reachable from the OpenCVE host; no public exploit code has been identified at time of analysis. The issue is fixed in OpenCVE 3.0.0.
Stack exhaustion in OpenImageIO's FITS image reader (all releases prior to 3.0.20.0, 3.1.15.0, and 3.2.0.3-beta1) lets a crafted FITS stream crash a parsing process by driving FitsInput::read_fits_header() into unbounded recursion. The trigger is a stream of consecutive 2880-byte header blocks that never contain the mandatory END keyword, so each block causes another self-call until the application stack is consumed; remote, unauthenticated attackers can deliver the payload but successful exploitation requires the victim to open or ingest the file (UI:R), and the resulting impact is availability-only with no code execution or data disclosure. No public exploit code or confirmed active exploitation has been identified at time of analysis, and vendor-released fixes are available.
Virtual host isolation bypass in IBM WebSphere Application Server and WebSphere Application Server Liberty, as bundled in IBM CICS TX Advanced 10.1, allows remote requests addressed to one configured virtual host to reach resources belonging to another virtual host on the same instance. The vendor assigns CVSS 3.1 6.5 (AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N, CWE-284); our independent assessment keeps the same unauthenticated, low-confidentiality/low-integrity, no-availability profile but raises attack complexity to AC:H because the weakness is only meaningful when an instance is configured to serve multiple distinct virtual hosts, while a default single-virtual-host deployment offers little to bypass. No public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis, and a vendor patch is available per the IBM advisory.
HTTP request smuggling in IBM WebSphere Application Server 8.5 can allow unauthenticated remote attackers to desynchronize request processing when the server sits behind or in front of an HTTP intermediary with mismatched Content-Length parsing, potentially leading to credential theft, cache poisoning, or request hijacking. The flaw (CWE-444) rates 6.5 (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:L/A:N) and requires a specific proxy-layer mismatch and concurrent traffic; no public exploit identified at time of analysis. Vendor patch is available and the risk is assessed as moderate and non-urgent.
HTTP request smuggling in Netty's HTTP/1 decoder lets remote, unauthenticated attackers bypass front-end security controls and reach resources they should not access by sending requests with malformed Transfer-Encoding headers. Netty is a widely embedded Java networking library, so the practical reach spans Red Hat products built on it - AMQ Broker 7, AMQ Clients, Red Hat Build of Keycloak, Red Hat Build of Quarkus, Red Hat Build of Apache Camel (Spring Boot 4 and Quarkus 3), Red Hat Build of Debezium 3, Red Hat Build of Apicurio Registry 3, Red Hat Data Grid 8 and Red Hat Fuse 7 - as well as JBoss EAP 7/8 and Red Hat Single Sign-On 7 per the associated advisory tags. CVSS 3.1 rates it 6.5 (AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N), consistent with smuggling-style control bypass rather than full compromise; no public exploit code or confirmed active exploitation was identified at time of analysis, though the flaw is trivially reachable on any exposed HTTP endpoint.
HTTP request smuggling and header injection in Netty's HTTP/2 stack (as bundled in multiple Red Hat products) allow remote unauthenticated attackers to bypass security controls and manipulate requests when HTTP/2 traffic is translated to HTTP/1.1. Affected deployments include Red Hat AMQ Broker 7, Camel, Keycloak, Quarkus, Data Grid, Fuse, EAP, Apicurio Registry, and Debezium. Assessed as high severity (CVSS 3.1 base 7.5, AV:N/AC:H/PR:N/UI:N/S:C/C:L/I:H/A:N); exploitation requires the specific topology of an HTTP/2-to-HTTP/1.1 translation boundary, and no public exploit was identified at time of analysis.
This is a real but moderate-priority vulnerability whose severity depends almost entirely on deployment topology rather than the raw 6.5 CVSS score. Netty's HTTP decoder lets split or duplicated Transfer-Encoding header fields slip past its final-chunked validation, so an unauthenticated, remote attacker who can send a raw crafted request to a Netty listener that sits in an HTTP chain with a second HTTP parser (reverse proxy, load balancer, CDN, WAF, or upstream/downstream server) can desynchronize the two parsers and smuggle a request (CWE-444). Impact is confined to low confidentiality and integrity loss (C:L/I:L) with no availability impact, no authentication (PR:N) or user interaction (UI:N) is required, and no public exploit identified at time of analysis.
A signed/unsigned type mismatch in Netty's Memcache binary protocol codec desynchronizes frame boundaries, allowing an unauthenticated, network-positioned attacker to smuggle responses across a shared or proxied memcache connection (CVSS 3.1 base 6.5, Medium). The flaw is reachable only when the Memcache binary codec is explicitly wired into a network-exposed Netty pipeline - a non-default configuration - and the affected code ships inside Red Hat Fuse 7, JBoss Enterprise Application Platform 7, Single Sign-On 7, and the Red Hat Build of Apache Camel for Spring Boot 4. No public exploit code has been identified at time of analysis, and there is no confirmation of active exploitation (not in CISA KEV); impact is limited to confidentiality and integrity (C:L/I:L) with no availability effect.
An out-of-bounds write in osmo-ggsn 1.14.0 allows a remote, unauthenticated sender of GTP signalling to corrupt process memory by delivering a PDP context whose GSN-Address sub-field is malformed; the decoder writes length-driven data without adequately bounding it to the destination buffer. The practical impact is availability-oriented - the assessed vector (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:H) indicates a low-complexity remote trigger that can crash or destabilise the GGSN, with a limited integrity effect, rather than a confirmed confidentiality breach. No public exploit code identified at time of analysis; EPSS is low (0.14%, 4th percentile) and CISA SSVC records Exploitation as 'none', though Automatable is 'yes'. Vendor/reference commit-level fix code is available upstream.
Unauthenticated path traversal in the Synology DiskStation Manager (DSM) Desktop API allows remote attackers to write a limited set of files and cause a limited denial-of-service condition on DSM instances older than 7.2.1-69057-10, 7.2.2-72806-7, or 7.3.2-86009-2. Because the CVSS vector is AV:N/AC:L/PR:N/UI:N, exploitation requires only network reachability to the DSM web management interface - no credentials or user interaction - though the impact is deliberately bounded: there is no confidentiality loss (C:N), with only low integrity and low availability effects from constrained file writes and partial service disruption. No public exploit code or confirmed in-the-wild exploitation has been identified at time of analysis, and an EPSS score was not available in the supplied intelligence.
Information disclosure in Synology DiskStation Manager (DSM) allows remote authenticated users to obtain sensitive information by directly requesting the Wallpaper Path resource, bypassing intended navigation controls. Affected versions are DSM before 7.2.1-69057-10, 7.2.2-72806-7, and 7.3.2-86009-2. Exploitation requires a valid DSM account (PR:L) and network access to the DSM web interface; no public exploit identified at time of analysis. Vendor patches are available in the listed fixed builds.
An authenticated SQL injection in the Sharing API of Synology DiskStation Manager (DSM) allows remote users with valid - even low-privilege - DSM credentials to read arbitrary sharing files they should not be able to access. The flaw (CWE-89) affects DSM builds prior to 7.2.1-69057-10, 7.2.2-72806-7, and 7.3.2-86009-2, and is constrained by its credential prerequisite: no user interaction is required, but a logged-in session is, which limits it to a confidentiality-only impact (CVSS 6.5, C:H/I:N/A:N) rather than full system compromise. At time of analysis there is no CISA KEV listing and no public exploit code identified, so the practical risk is a bounded data-disclosure issue for organizations that expose the DSM Sharing API to untrusted or broad authenticated user populations.
Time-based blind SQL injection in the WordPress plugin Photo Gallery by 10Web - Mobile-Friendly Image Gallery (all versions up to and including 1.8.44) lets an attacker who holds an Author-level account or higher smuggle SQL into the album_id shortcode attribute and read data that the WordPress database user can access. Because the crafted attribute is stored inside a published post, the injected query fires for every visitor who renders that post, and the unsanitized value is placed on both sides of a UNION statement, which can roughly double the timing delay an attacker measures. There is no CISA KEV entry and no public exploit identified at time of analysis; the CVSS 3.1 base score is 6.5 (AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N), and practical exploitation is constrained by the required authenticated Author account and by the slow, blind extraction technique.
Authenticated SQL injection in the WP Inventory Manager plugin for WordPress (versions up to and including 2.5.1) allows users with Contributor-level access or above to read sensitive data from the site database by injecting SQL through the 'where' attribute of the [wpinventory] shortcode. The root cause is in WPIMItem::get_all()/parse_custom_where(), which does not use $wpdb->prepare() or a field whitelist before concatenating the attacker-influenced string into a raw SELECT executed via $wpdb->get_results(); the assessed impact is confidentiality-only (CVSS 6.5, C:H/I:N/A:N). No public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis, and exposure is limited to sites that run the affected plugin, render the shortcode, and grant Contributor-or-above accounts to untrusted users.
Authenticated attackers with Author-level access or higher on WordPress sites running Download Manager (Codename065) versions up to 3.3.68 can exploit an insecure direct object reference to duplicate arbitrary wpdmpro packages they do not own, including those belonging to administrators. The flaw lies in the duplicate() handler hooked on admin_init, which checks only a generic 'edit_posts' capability and a plugin-wide static nonce, omitting object-level authorization. This allows the attacker to clone a target package, inherit its protected file references and access restrictions, then edit the clone to strip those restrictions and download the previously protected files. No public exploit identified at time of analysis; exploitation requires an authenticated account and is limited to environments that grant untrusted users author or contributor roles.
SQL injection in the InfiniteWP Client plugin for WordPress (versions up to and including 1.13.9) allows an attacker who already holds WordPress administrator-level access to read arbitrary data from the site's database through the plugin's get_comments action. Exploitation requires the attacker to first register their own public key via the plugin's add_site action using the plugin's WP-admin-generated activation_key, then send cryptographically signed get_comments requests whose JSON body (read from php://input, so wp_magic_quotes() never sanitizes it) carries crafted array-key names that are concatenated into an unparameterised IN(...) SQL clause. Wordfence reported the issue and an upstream fix is referenced in the plugin repository, and no public exploit code has been identified at time of analysis; the practical impact is bounded by the high privilege prerequisite, since an administrator already has broad control over the site, making the realistic risk lower than the 6.5 CVSS score or a bare "SQLi" label implies.
Metadata disclosure in Nextcloud Server's Photos app smart albums allows an authenticated user to enumerate file names, IDs, and other metadata from a sharer's folders that were never intended to be included in the shared album. This moderate, tightly scoped authorization flaw (CWE-284) occurs because the smart album builds its file listing using the viewing user's photosSourceFolders search configuration instead of the album owner's configuration; an attacker who has been granted a filter-based smart album share can modify their own folder settings to widen the search scope. No public exploit code is identified at time of analysis, and the independent assessment limits impact to metadata (not file content) within the filter-reachable scope; it cannot reach files of users who have not shared such an album.
Authorization bypass in the Nextcloud team folders app (formerly group folders), when used together with the workspace app, allows authenticated delegated administrators to reach folders outside the scope they were granted. Affected are Nextcloud deployments where organizations delegate limited team-folder administration strictly through API/REST; the attacker must already hold advanced permissions on at least some folders (CVSS 3.1 AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:N, score 6.5), so unauthenticated users and ordinary accounts cannot exploit it. Confidentiality and integrity of folders outside the delegated scope are both compromised, availability is not affected, and no public exploit code has been identified at time of analysis - the CVE is not listed in CISA KEV, and the report originates from HackerOne (report 3674940).
Stored cross-site scripting in the Popup Maker plugin for WordPress versions up to and including 1.24.0 allows authenticated attackers with contributor-level access or higher to inject persistent JavaScript into post titles, which executes in the browsers of users viewing affected admin pages. The vulnerability has a CVSS score of 6.4 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:N) and is reported by Wordfence; no public exploit code or active exploitation has been identified at time of analysis. The issue is patched in version 1.25.0.
Server-side request forgery in the IBM i FTP server (releases 7.3, 7.4, 7.5, and 7.6) lets an authenticated FTP user bypass security restrictions and pivot to internal network services by supplying crafted PORT and EPRT commands. The flaw is an improper-validation issue (CWE-918) with a scope change because the FTP component directs connections to services outside its own security boundary; scoring is CVSS 6.4 (AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:N), so exploitation requires valid FTP credentials and network reachability of the FTP service, and impact is limited to confidentiality and integrity of whatever internal services the IBM i host can reach. Per the independent assessment this is a genuine but moderate-priority issue rather than an emergency, and no public exploit identified at time of analysis; the vendor has published a fix.
Stored cross-site scripting in the WordPress plugin RT Mega Menu - Mega Menu Builder for Elementor & Gutenberg (all versions through 1.5.1) allows an attacker with an authenticated Subscriber-level account or higher to inject arbitrary JavaScript through the 'css[left]' mega-menu styling parameter, which is persisted and then executed in the browser of every visitor who loads a page rendering the poisoned menu. The assessed severity is moderate (CVSS 6.4, CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:N), with the low privilege bar rather than the payload itself being the main limiting factor; no public exploit code and no confirmed active exploitation were identified at time of analysis, so priority should be driven by whether the site allows self-registration or hosts untrusted low-privileged users.
Authenticated stored cross-site scripting (XSS) in the Brizy Page Builder plugin for WordPress versions up to and including 2.8.14 allows attackers with Contributor-level access or higher to inject arbitrary web scripts via the rootAttributes parameter, which execute when a victim views the affected page. The vulnerability requires authentication and a subsequent page view by another user, and no public exploit has been identified at time of analysis. The CVSS score is 6.4 with a scope change, reflecting limited confidentiality and integrity impact.
Stored cross-site scripting in the Kubio AI Page Builder plugin for WordPress (all versions up to and including 2.8.4) lets authenticated users with contributor-level access or higher embed HTML-entity-encoded script markup inside the kubio/copyright block, which the plugin subsequently decodes back into live HTML and renders to anyone who views the page. Because the payload survives both the WordPress core kses-on-save filter and the plugin's own wp_kses_post() call, the injected script executes in the browser of an admin, editor, or visitor who loads the affected content; the attacker needs no further interaction after saving the post. Rated CVSS 6.4 (AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:N) consistent with an authenticated-authoring prerequisite and cross-user impact; no public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis.
Authenticated attackers with contributor-level access can inject persistent JavaScript into WordPress sites running Strong Testimonials 3.3.8 or earlier through the 'lightbox_class' shortcode attribute, due to insufficient input sanitization and output escaping. Exploitation requires that the targeted testimonial view has at least one published testimonial with a featured image and the lightbox wrapper enabled, because the vulnerable code path is only reached when a thumbnail is rendered; the stored script then executes automatically when any user views the injected page. No public exploit code or active exploitation has been identified, and the CVSS score of 6.4 reflects a scope change with low confidentiality and integrity impact.
Stored cross-site scripting in the WordPress plugin VK All in One Expansion Unit (all versions through 9.118.0) allows an authenticated attacker holding contributor-level privileges or higher to plant arbitrary JavaScript in the 'vkExUnit_cta_img_position' post meta, which is then rendered verbatim into the Call-To-Action block output and executes in the browser of any visitor or administrator who later views the affected page. The save-side sanitize_text_field() callback fails to strip double-quote characters and on* event-handler attributes, and the plugin's Vk_Call_To_Action::safe_kses_post() output filter only rewrites disallowed iframe elements while passing everything else through unchanged, so the injected markup survives intact. No public exploit code was identified at time of analysis and the issue is not confirmed actively exploited (CISA KEV), but exploitation requires only a low-privilege contributor account, making it most impactful on multi-author sites or blogs that permit open or untrusted user registration; the assessed CVSS 3.1 base score is 6.4 (AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:N).
Authenticated stored cross-site scripting in the Custom Twitter Feeds (A Tweets Widget or X Feed Widget) WordPress plugin affects all versions up to and including 2.8.0, allowing an attacker holding a contributor-level account or higher to embed arbitrary script in the 'buttoncolor' attribute of the plugin's legacy shortcode. The injected markup is stored and executes in the browser of any user - including an administrator - who later views the affected page, with the cross-boundary scope (S:C) reflecting that the payload runs in the context of the hosting site rather than the plugin itself. Exploitation is conditional: the legacy code path must actually be active, and per the vendor analysis this occurs by default on common installs where the admin configured access tokens together with zero or multiple legacy feeds, which sets the ctf_statuses 'support_legacy_shortcode' option to boolean true. At time of analysis there was no public exploit identified and the issue is not recorded as confirmed actively exploited (CISA KEV), so the practical risk is a post-authentication content-injection scenario rather than an unauthenticated remote compromise.
Limited server-side request forgery in the Auto Upload Images WordPress plugin (all versions up to and including 3.3.2) lets an authenticated user with contributor-level access or higher make the affected site issue outbound web requests to arbitrary destinations. The plugin's downloadImage routine uses wp_remote_get() instead of wp_safe_remote_get(), and its validate() method only rejects URLs whose host matches the site's own hostname, so requests to loopback, private, and link-local targets such as 127.0.0.1, 10.0.0.0/8, or 169.254.169.254 are not blocked. Exploitation is triggered simply by embedding a crafted <img> tag in post content and submitting it for processing; no KEV entry, public exploit code, or other active-exploitation signal was identified at time of analysis, and the practical impact is bounded because the SSRF is a GET-based image fetch with low confidentiality impact and a contributor authentication prerequisite.
Stored cross-site scripting in the Magazine Blocks WordPress plugin (wpblockart) through version 1.8.6 lets an authenticated user with contributor-level access or higher store arbitrary JavaScript that executes in the browser of anyone viewing the affected page. The flaw sits in the News Ticker block, where the block's clientId attribute is concatenated into an HTML class attribute inside NewsTicker::render() without esc_attr(), so the payload is persisted in post content rather than reflected. Exploitation requires a valid low-privilege account plus access to the block editor, which limits exposure to multi-author sites with open contributor registration or low-trust editorial roles; no public exploit code, EPSS signal, or CISA KEV entry was identified at time of analysis, and this is a genuine but moderate-priority issue rather than an inflated high-severity finding.
Stored cross-site scripting in the WordPress plugin RT Mega Menu - Mega Menu Builder for Elementor & Gutenberg (all versions up to and including 1.5.2) allows authenticated attackers holding contributor-level or higher privileges to plant a tag-less script payload in a block's 'pointer_menu_item' attribute, which then executes in the browser of any administrator, editor, or visitor who opens the injected page. Because the payload contains no HTML tags it slips past the wp_kses_post filtering applied on save and survives intact inside the block comment's JSON attributes, so the plugin's navigation walker renders it unescaped. Vendor CVSS 6.4 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:N); no public exploit identified at time of analysis and no confirmed active exploitation (not in CISA KEV); the contributor-level authentication gate limits real-world exposure to sites that grant untrusted users authoring accounts.
Memory-exhaustion denial of service in ImageMagick before 7.1.2-31 stems from the Ultra HDR (UHDR) encoder omitting resource-policy checks while allocating the pixel buffer, so a crafted UHDR image can push allocations past the memory/area/disk limits an operator configured in policy.xml. The flaw is remotely reachable without authentication (CVSS 4.0 vector PR:N/UI:N) in any application that routes untrusted images through ImageMagick's UHDR encoding path, but the independent assessment rates it low priority: AC:H and AT:P reflect a niche code path and a crafted image that must reliably defeat the policy logic, and impact is confined to availability (VC:N/VI:N/VA:L) with no confidentiality, integrity, or code-execution effect. Vendor patch 7.1.2-31 is available; no public exploit code and no confirmed active exploitation were identified at time of analysis.
A division-by-zero in ImageMagick's FLIF encoder allows unauthenticated remote attackers to crash the encoder process by supplying an image whose ticks-per-second field yields a zero divisor, causing a denial of service. Affected versions are ImageMagick before 7.1.2-31 (7.x) and before 6.9.13-56 (6.x); exploitation requires that a target application encode attacker-influenced data to FLIF, a relatively uncommon output format, and no public exploit code or active exploitation has been identified at time of analysis.
Sampling-parameter validation in vLLM before 0.29.0 checks the allowed_token_ids allowlist against the tokenizer vocabulary size instead of the model's output logits width, so a remote client calling the OpenAI-compatible inference API can supply token IDs that are valid for the tokenizer but out of range for the model output (the advisory's cited case is a multimodal model with 128257 tokenizer tokens versus 128256 output logits). Those out-of-range IDs pass validation and reach the GPU sampling path, where LogitBiasState's Triton logit-bias kernel indexes past the valid logits row and corrupts shared GPU logits state, causing concurrent victim requests in the same batch to sample tokens outside their own allowlists. The impact is limited to sampling-constraint integrity (no data disclosure, crash, or code execution) in shared or multi-tenant deployments, and it is inert on models whose tokenizer and output vocabularies are the same size; there is no public exploit identified at time of analysis and no confirmed active exploitation (not listed in CISA KEV).
Out-of-bounds writes in the Triton bincount sampling kernel of vLLM through 0.29.0 let unauthenticated remote clients corrupt a neighboring request's repetition-penalty state by submitting multimodal audio input that contains a token ID equal to the model's vocabulary size. The root cause is an unvalidated array index (CWE-129): prompt and output token IDs are used to index the penalty prompt-presence bitset and output bin counts without any bounds check against vocab_size, so a one-past-the-end token spills into the adjacent request's mask/count rows. Impact is integrity-only and confined to altered repetition-penalty behavior for concurrently served requests - there is no code execution, data disclosure, or denial of service - and exploitation requires a CUDA GPU with compute capability >= 8.0, a multimodal model accepting audio tokens, and concurrent traffic in a shared, multi-tenant deployment; no public exploit code and no confirmed active exploitation have been identified at time of analysis.
Path traversal in the ASReportService component of HCL AppScan 360° allows an authenticated user to supply manipulated file-path input and reach files outside the intended report directory, exposing the surrounding file-system structure or permitting modification of files within the application's directory scope. The flaw is network-reachable but requires a valid account and access to the ASReportService interface (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:L), which keeps practical risk moderate and well below the 6.3 base score. There is no public exploit identified at time of analysis and no confirmed active exploitation.
Cross-user lock manipulation in the Nextcloud Files Lock (files_lock) app allows any authenticated user to lock or unlock files they do not own by supplying the absolute WebDAV path of another user's file, because the DAV plugin resolves the request URI without confirming that the path belongs to the session user. An attacker who can enumerate or guess a victim's path can block editor saves and PUT/MOVE/DELETE writes, and can also have the application hand back the victim's lock token, which enables removal of token-based client locks. Rated 6.3 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:L) in the authoritative assessment, it requires a valid account and an enabled files_lock app on a multi-user instance; impact is limited to denial-of-write and lock-token disclosure with no file content read/write or code execution, and no public exploit code or confirmed active exploitation was identified at time of analysis.