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.
Memory exhaustion and denial of service in FluidSynth 2.5.0 through 2.5.5 allows an attacker to crash any application that parses an attacker-supplied SoundFont 2 file. The SF2 parser derives the DMOD modulator count as chunk.size / SF_MOD_SIZE - 1 without first rejecting chunks smaller than a single record, so a zero-sized DMOD chunk underflows the unsigned subtraction to UINT_MAX and the loader attempts to allocate billions of SFMod structures until process memory is exhausted. Rated CVSS 6.2 (AV:L/AC:L/PR:N/UI:N, availability-only), the issue requires only that a crafted .sf2 file be loaded locally (no privileges and no user interaction beyond opening the file), no public exploit code or confirmed active exploitation has been identified at time of analysis, and the vendor-released patch is version 2.5.6.
Cross-site scripting in Moonshot AI's Kimi web application (kimi.com, affected as of 2026-07-18) lets a remote, unauthenticated attacker publish a crafted HTML artifact through the public Share view so that the artifact Preview rendering path executes attacker-supplied JavaScript in the Kimi origin. The flaw is exploitable only against a victim who opens the attacker's share link (CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N, CVSS 6.1) - there is no zero-click path, and impact is confined to the victim's browser session in the Kimi origin (S:C with limited confidentiality and integrity effects). The references include a public GitHub repository (MGTx2/CVE-2026-79294) that appears to document the issue, so proof-of-concept material may be publicly accessible; there is no CISA KEV entry and no confirmation of active exploitation at time of analysis. Note that although the CVE description says 'execute arbitrary code,' the assessed impact is arbitrary client-side JavaScript execution, not server-side code execution.
Blind server-side request forgery in the Nextcloud Circles (Teams) app lets any unauthenticated remote user force the Nextcloud Server to issue an outbound GET request to an attacker-chosen internal or private address through the public endpoints POST /apps/circles/event/ and POST /apps/circles/incoming/. The flaw sits in remote-instance signature verification, which fetches the attacker-supplied keyId URL before trust in the remote instance is established and explicitly permits local and private address ranges, bypassing Nextcloud's core SSRF protections. Because the internal response body is never returned to the requester, the practical impact is internal-network reconnaissance - an attacker can map which services are reachable - rather than data theft or code execution; no public exploit code has been identified at time of analysis and the issue is not a confirmed actively exploited (CISA KEV) vulnerability. The independent assessment rates this CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N, treating it as a moderate-severity reconnaissance primitive.
Cross-site scripting in the hosted Microsoft Azure Portal (CVE-2026-83946) allows an unauthenticated attacker to have malicious script rendered inside the privileged portal origin, escaping into a trusted administrative context once an authenticated portal user (typically an administrator) loads or clicks attacker-controlled content. The vendor CVSS 3.1 vector AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:L/A:N rates this 8.2 (High), driven chiefly by the scope change (S:C) and high confidentiality impact (C:H) - meaning injected script can read data across the portal origin - with the required victim interaction (UI:R) as the main limiting factor. No CISA KEV listing and no public exploit code was identified at time of analysis, and because Azure Portal is a Microsoft-operated service, the issue was remediated server-side, so no customer patch or upgrade exists or is needed.
Cross-site scripting in the IBM Common Licensing Agent and ART Web UI (9.0, 9.0.0.1, 9.0.0.2) lets an unauthenticated attacker embed arbitrary JavaScript into the interface, which executes in the browser of any user who is lured into rendering the crafted content; because the script runs inside a trusted, authenticated licensing session, the realistic outcome is theft of that session's credentials or manipulation of the displayed licensing data. The CVSS vector (AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N, score 6.1) confirms no attacker authentication is needed but that exploitation depends on victim interaction, and the changed scope indicates the injected payload can affect resources beyond the vulnerable component. There is no public exploit code identified at time of analysis and the issue is not confirmed actively exploited (CISA KEV); the vendor has published a fix, and the primary limiting factor remains the narrow footprint of the licensing component and the requirement for a reachable Web UI plus a legitimate user to trigger the payload.
Reflected cross-site scripting in IBM Financial Transaction Manager for SWIFT Services for Multiplatforms (versions 3.2.4.0 through 3.2.4.16) lets an unauthenticated attacker inject arbitrary JavaScript into the Web UI, which then executes in the browser of any authenticated user who loads the crafted content. Because the payload runs inside a trusted, already-authenticated session, the realistic impact is disclosure of that user's credentials or session data rather than server compromise; exploitation is constrained by the mandatory requirement that a victim interact with attacker-supplied content (CVSS 6.1, AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N/). No public exploit code or confirmed active exploitation was identified at the time of analysis, and the vendor has published remediation guidance.
Reflected cross-site scripting (XSS) in the Qi Addons For Elementor plugin for WordPress (versions up to and including 1.11) allows unauthenticated remote attackers to inject arbitrary web scripts via the 's' search parameter, but exploitation is highly conditional: it requires the Table of Contents widget to be placed on a template that renders on the WordPress search-results page (e.g., sitewide header or footer) with the 'Limit ToC to Main Page Content' option left at its default value of No, and it requires victim interaction (opening a crafted search URL). The vulnerability has a CVSS 3.1 base score of 6.1 (AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N), and no public exploit identified at time of analysis.
Reflected cross-site scripting in the ShopLentor (WooLentor Addons) WordPress plugin up to and including version 3.5.1 lets unauthenticated attackers inject arbitrary JavaScript into a page by abusing the name of a query-string parameter, which the WL: Product Horizontal Filter widget reflects unescaped into an option element value attribute. Exploitation requires the victim to click an attacker-supplied link that carries the payload in the parameter name (UI:R) and only works on published pages where that widget has actually been placed, so the result is non-persistent script execution scoped to the WordPress origin; the higher-value outcome, abuse of an authenticated administrator session, requires luring a logged-in admin to the crafted URL. No public exploit code or confirmed in-the-wild exploitation was identified at time of analysis, and the issue is assessed as a genuine but moderate-severity problem (CVSS 6.1, scope-changed reflected XSS) rather than a mass-exploitation candidate.
Cross-site scripting in IBM Common Licensing Agent 9.0 through 9.0.0.2 and ART 9.0 through 9.0.0.2 lets remote, unauthenticated attackers (CVSS PR:N) embed arbitrary JavaScript in the licensing Web UI, which executes in the browser of a victim who interacts with the crafted content (UI:R), potentially disclosing credentials or session data within a trusted session (CVSS 6.1, S:C/C:L/I:L/A:N). Only these specific Agent and ART builds are affected, and because the component is a backend licensing interface rather than a public-facing web application, opportunistic drive-by exploitation is limited; no public exploit identified at time of analysis. Successful exploitation still depends on a logged-in operator or administrator viewing attacker-influenced content, which is the primary limiting factor.
Cross-site scripting and JSON injection in the Quarkus Qute template engine stem from the {#eval} section helper failing to pass the parent template's content type to the sub-template it evaluates, so contextual escaping is skipped and untrusted data is emitted as raw HTML or JSON. The flaw reaches Red Hat Build of Quarkus and downstream Red Hat offerings that bundle Qute (Red Hat Build of Apache Camel 4 for Quarkus 3, Red Hat Build of Keycloak, Red Hat Build of Apicurio Registry 3 and Red Hat Fuse 7), but only where an application passes attacker-influenced data into {#eval}. The assessed vector is CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N (6.1): payload submission needs no authentication (PR:N), yet a victim must render the affected page for script execution to occur, and impact is confined to that victim's browser context (C:L/I:L/A:N); no public exploit identified at time of analysis.
Cross-site scripting in HCL BigFix Service Management lets an attacker who already holds a valid low-privilege account (CVSS PR:L) inject unsanitized script into content that executes in another user's browser when that user views it (UI:R), enabling session hijacking, account takeover, and unauthorized actions performed as the victim. Impact is greatest when the viewing victim is a higher-privileged user such as an agent or administrator, because the injected script then runs with that user's session context. No public exploit identified at time of analysis, and because the affected component is a niche enterprise ITSM module rather than broadly deployed software, this remains a moderate-priority issue.
Reflected cross-site scripting in Semantic MediaWiki 7.0.0 through 7.1.0 lets an unauthenticated attacker execute arbitrary script in a victim's browser on the wiki origin by tricking them into opening a crafted Special:Ask link carrying a forged cursor pagination token. The cursor parameter added in 7.0.0 is an unsigned base64url-encoded JSON blob, so its sort_prop and sort_order fields are fully attacker-controlled; when they mismatch the request's sort=/order= values, the resulting error string is interpolated into Query::addErrors() and rendered as raw HTML by Html::errorBox(), bypassing the message-layer sanitization that normally protects SMW error output. Exploitation requires victim interaction (clicking the link) and is aided by the absence of a Content-Security-Policy on responses, which permits inline script to run; there is no confirmed active exploitation and no public exploit code identified at time of analysis, though the maintainer advisory documents a browser-confirmed reproduction, and the CVSS 3.1 score is 6.1 with a scope change. Upgrading to 7.2.0 removes the vulnerability.
HTML/CSS injection in Open edX Platform discussion-notification emails allows an authenticated, enrolled student to embed CSS-capable markup in a discussion post title, which is then interpolated without sanitization into digest and batched-email notification content and rendered as trusted HTML; when another learner opens that email in a client that renders CSS, the attacker gains email-open tracking, content spoofing, and a credible phishing vector. The issue affects releases from Redwood up to (but not including) Ulmo and prior to Verawood.1, and is fixed in Ulmo and Verawood.1. Exploitation requires course enrollment, digest or batched (non-immediate) notification delivery, and victim interaction with the email, and no public exploit has been identified at time of analysis.
Heap out-of-bounds read in OpenImageIO's IFF image reader (src/iff.imageio/iffinput.cpp, IffInput::readimg()) affects all releases prior to 3.0.20.0, 3.1.15.0, and 3.2.0.3-beta1, allowing a caller that opens a crafted uncompressed 16-bit IFF file containing a z-buffer channel to trigger a crash or leak adjacent heap contents into the decoded image output. The flaw is an allocation/stride mismatch: the temporary scanline buffer is sized from m_header.rgba_count while the subsequent memcpy uses m_header.pixel_bytes(), whose stride also counts z-buffer bytes, producing an oversized copy. Exploitation requires local, user-initiated parsing of the malicious file (CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:L/I:N/A:H, 6.1) with no privileges needed but no remote zero-interaction path; no public exploit code and no confirmed active exploitation were identified at time of analysis, and EPSS/KEV data were not provided. Vendor-released patches exist in versions 3.0.20.0, 3.1.15.0, and 3.2.0.3-beta1.
Stack-based buffer overflow in OpenImageIO's Cineon image reader (CineonInput::open()) lets a crafted .cin file that declares more than the format's maximum of eight channels overwrite stack pointers and corrupt adjacent memory, crashing the application that processes it. Exploitation requires the victim or an automated pipeline to open the malicious file (CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:L/A:H, 6.1), so it cannot be triggered remotely without inducing file ingestion; the assessed impact is denial of service with only low integrity risk and no confidentiality loss. No public exploit identified at time of analysis and the issue is not in CISA KEV; vendor fixes shipped in OpenImageIO 3.0.20.0, 3.1.15.0, and 3.2.0.3-beta1.
Reflected cross-site scripting in Semantic MediaWiki 7.1.0 and earlier lets an unauthenticated remote attacker inject arbitrary markup into the query debug output rendered on Special:Ask, executing JavaScript in the browser of a victim who follows a crafted link. No account, privilege, or attacker-created wiki content is required: the predefined text-type properties such as Text (_txt) that ship on every default install satisfy the property-type condition the payload needs, so a URL like Special:Ask?q=[[Text::<script>...</script>]]&debug=1 is sufficient. The CVSS:3.1 vector AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N yields a 6.1 base score, reflecting a scope-changing, low-impact script injection that depends on user interaction and is therefore not self-propagating or wormable; no public exploit code has been identified at time of analysis, and the issue is not confirmed actively exploited (no CISA KEV listing). Vendor-released patch: 7.2.0.
Reflected cross-site scripting in Semantic MediaWiki 7.1.0 and earlier allows unauthenticated attackers to inject arbitrary HTML markup into the Special:Ask table output by abusing the `sep` (separator) parameter, which was inserted verbatim into the HTML that joins a table cell's values. Because the same unsanitized table HTML is produced for the raw request output (request_type=raw), no authentication is needed to reach the injection, though exploitation is limited by user interaction: a victim must be induced to load a crafted query URL for the payload to execute in their browser. The issue is correctly rated CVSS 6.1 (AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N) as a moderate-severity reflected XSS; no public exploit code and no confirmed active exploitation were identified at time of analysis, and the vendor has shipped a fix in version 7.2.0.
Reflected cross-site scripting in Semantic MediaWiki 7.1.0 and earlier allows an attacker to inject script into the Special:SearchByProperty page through the `property` and `value` query parameters, executing arbitrary JavaScript in the browser of any wiki user lured to a crafted URL. Rated CVSS 6.1 (AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N), the flaw needs no authentication to construct or deliver the malicious link but does require victim interaction (UI:R), and impact is confined to the victim's browser session with no server-side code execution. No public exploit code or confirmed active exploitation (CISA KEV) was identified at time of analysis; the vendor released a fix in SMW 7.2.0 that escapes both the reflected form value and the derived error messages.
Reflected cross-site scripting (CWE-79) in Semantic MediaWiki 7.1.0 and earlier allows a crafted URL to execute script in a victim's browser session; the flaw is reached through the Special:Ask result printer when it is invoked with headers=plain and an attacker-controlled mainlabel value, which is written into the table header (<th>) through a raw HTML path instead of being escaped for the output context. The vulnerability is rated CVSS 6.1 (AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N), meaning no authentication is needed to craft the link, but exploitation depends on the victim opening the malicious URL (user interaction is required) and impact scales with the victim's wiki privileges. No public exploit code and no confirmed active exploitation were identified at time of analysis; the issue is fixed in Semantic MediaWiki 7.2.0, so deployments on 7.1.0 or earlier that expose Special:Ask queries are the realistic exposure.