Improper boundary validation of metadata slot values in luksmeta allows an attacker who can supply or modify a LUKS disk header to redirect wipe or load operations outside the intended metadata area, overwriting or zeroing encrypted payload data and causing persistent data loss and corrupted storage. The attack is confined to a local vector and requires high privileges - root-level write access to the raw block device, or inducing an administrator to process an attacker-supplied untrusted disk image whose header carries out-of-range slot values together with valid checksums. The impact is limited to integrity and availability, with no information disclosure; no public exploit code and no confirmed active exploitation were identified at time of analysis, and the moderate CVSS of 6.0 reflects the substantial privilege barrier rather than the severity of the data-loss outcome.
Repeated state-notification requests sent to a network-reachable rpc.statd service in nfs-utils cause memory allocated during callback processing to never be released, producing continuous memory growth that can degrade or crash the service and exhaust host resources. The flaw is remotely triggerable without authentication (CVSS:3.1/AV:N/AC:H/PR:N/UI:N) but exploitation is not trivial: a single request is insufficient because the attacker must sustain a stream of requests long enough to exhaust memory, which is why the vector carries AC:H and the impact is availability-only (C:N/I:N/A:H, base score 5.9). Exposure is limited in practice because rpc.statd only runs when NFSv3 locking services are active and is normally bound to internal or trusted networks rather than the public internet, so firewalled or default deployments are not reachable. No public exploit code has been identified at time of analysis and the issue is not in CISA KEV; this is a genuine but moderate-priority availability issue rather than a top-tier emergency, and affected services can be recovered by a restart.
Exception-detail disclosure in the Wheels CFML framework (versions 3.0.0 through 4.1.1) lets a remote, unauthenticated client receive full stack/exception details when the non-default setting allowIPBasedDebugAccess=true is enabled. Because error and debug-display flags live in shared application scope, an administrator request that changes those settings can race with an unauthorized client's erroring request, so that client may receive tag context or a serialized exception and any sensitive data inside it - confidentiality only (CVSS 5.9, CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N). Practical risk is well below what the network vector suggests: exploitation requires both the explicitly enabled non-default configuration and a narrow timing window, the race does not bypass the production gate that protects developer tools, and there is no public exploit identified at time of analysis. The issue is fixed in 4.1.2.
Information disclosure in the Wheels CFML framework through 4.1.1 allows an unauthenticated remote attacker who wins a cold-start race to redirect the application's production error-email recipient to an address they control, because the default recipient is derived from the attacker-supplied Host header instead of a configured value. Once the attacker's request initializes the application, error emails silently flow to that address for the application's lifetime, disclosing error and request metadata and - on engines and versions where scope dumps render (Lucee 7 does not render them, limiting exposure to header fields) - configuration secrets and other users' session and cookie data. Exploitation is highly conditional: it requires no explicitly configured error recipient, a configured mail server, an upstream server that accepts an arbitrary Host value, and the attacker's request coinciding with the cold start; no public exploit identified at time of analysis, and the flaw is fixed in 4.1.2.
Information disclosure in Microsoft Authentication Library (MSAL) for Node.js can expose sensitive data to an untrusted origin due to improper origin validation (CWE-346). The assessed CVSS 3.1 vector (AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N) allows unauthenticated remote exploitation without user interaction, but high attack complexity means success depends on specific MSAL flows or cross-origin conditions rather than default behavior. No public exploit identified at time of analysis; a vendor patch is available through the Microsoft MSRC update guide.
Tandoor Recipes deployments prior to 2.6.10 can be driven into denial of service by an authenticated user who stores a catastrophic-backtracking regular expression in a shared-Space Automation rule, which the server then evaluates with Python's re.sub() during another user's recipe import through the api/recipe-from-source/ endpoint. Because the pattern is executed without complexity validation or any execution timeout, a single crafted rule can monopolize a Gunicorn worker until the request timeout elapses, exhausting available worker slots and making the application unavailable to legitimate users. Exploitation requires a valid authenticated account that belongs to a shared Space plus a second user importing a matching recipe (CVSS 5.7, CWE-1333, AV:N/AC:L/PR:L/UI:R with availability-only impact), no public exploit code has been identified at time of analysis, and the vendor shipped a fix in 2.6.10.
Processing a crafted MP4 file with GPAC MP4Box 2.4 can terminate the tool with SIGFPE, because the fraction-string parser gf_parse_lfrac() in src/utils/error.c divides by a zero denominator supplied by the file's metadata. Only the single MP4Box process crashes; there is no code execution and no confidentiality or integrity impact (CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H, 5.5). Exploitation is local and requires the victim or an automated pipeline to open the malicious file, EPSS is a negligible 0.16% (4th percentile), and there is no public exploit identified at time of analysis - an availability-only denial-of-service with LOW priority despite the media-parser context.
Heap memory disclosure in GStreamer gst-plugins-good before 1.28.2 occurs when the isomp4 qtdemux parser processes a crafted MP4/QuickTime 'uuid' atom and reads past the end of a heap-allocated buffer. The issue has a local attack vector with user interaction (CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:N/A:N), so a victim must load the malformed media file in a GStreamer-based application, and impact is limited to confidentiality with no code execution or crash claimed. No public exploit code or CISA KEV active exploitation is identified at time of analysis, and EPSS is very low at 0.15% (4th percentile).
IBM i releases 7.3, 7.4, 7.5, and 7.6 can disclose sensitive information to a local attacker who already holds an authenticated low-privileged account, because file handle resolution omits necessary subtree checks and therefore fails to enforce the directory hierarchy boundary. The vendor CVSS vector rates the issue at 5.4 (network-vector, PR:L), but the independent assessment re-scopes it to a local attack (AV:L) with confidentiality-only impact, meaning an attacker needs an existing foothold on the system rather than unauthenticated remote access. There is no public exploit code identified at time of analysis and the issue is not confirmed actively exploited (not in CISA KEV); EPSS data was not provided, so risk is best treated as moderate with a low likelihood of mass exploitation and a higher payoff for an insider or post-compromise attacker seeking data outside their authorized subtree.
Incorrect DBus policy in lomiri-download-manager, the download service shipped with Ubuntu Touch (UBports), allows an unprivileged local user to invoke privileged methods on the root-owned system bus service and execute arbitrary commands as root. The affected component is the UBports lomiri-download-manager package across all released versions (CPE cpe:2.3:a:ubports:lomiri-download-manager:*), and exploitation is strictly local: it requires the attacker to already hold an unprivileged account or in-app code execution on the device, with no user interaction needed. Vendor input rates a lower CVSS 3.1 base score of 5.5 (C:H/I:N/A:N, scope unchanged), while independent analysis assesses the correctly scoped vector as CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H, reflecting full root compromise as well as unauthenticated arbitrary file read as root; no public exploit code or confirmed active exploitation was identified at time of analysis.
Local denial of service in the nfs-utils fsidd daemon affects Red Hat Enterprise Linux 6, 7, 8, 9, and 10 as well as Red Hat OpenShift Container Platform 4; an authenticated local user with access to the fsidd Unix domain socket can send a malformed request that omits required arguments, triggering a CWE-617 assertion failure and crashing the daemon. The service remains unavailable for filesystem identifier lookups until restarted, giving this a CVSS 3.1 score of 5.5 (AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H). No public exploit code has been identified at time of analysis.
Stored cross-site scripting in the Events Made Easy WordPress plugin (all versions up to and including 3.2.22) lets unauthenticated attackers inject arbitrary script into the event[location_name] field, which then executes in the browser of any visitor who views the affected location map balloon. Per the CVSS 3.1 vector (AV:N/AC:H/PR:N/UI:N/S:C/C:L/I:L/A:N) the flaw is remotely reachable without credentials, but exploiting it requires an administrator to have explicitly enabled both the guest event submission feature (eme_fs[guest_submit]) and forced location creation (eme_fs[force_location_creation]) - both of which default to off - so default-configured sites are not exposed. There is no public exploit identified at time of analysis and the plugin is not in CISA KEV; the vendor addressed the issue in version 3.2.23.
Stored cross-site scripting in the WPeMatico RSS Feed Fetcher WordPress plugin (etruel) affects all versions up to and including 2.9 and allows an unauthenticated remote attacker to plant arbitrary web scripts inside imported posts. The attacker does not need a WordPress account: because the plugin strips WordPress's wp_filter_post_kses filter from content_save_pre immediately before calling wp_insert_post() whenever a campaign's last editor holds the unfiltered_html capability - a flag the plugin sets automatically for every admin-saved campaign - raw RSS markup taken from $item->get_content() is written verbatim into post_content. Exploitation therefore depends on an administrator having subscribed a campaign to an RSS feed the attacker publishes or compromises, and the injected script only fires when a visitor later opens the imported post; no public exploit code or confirmed active exploitation was identified at time of analysis, and the assessed CVSS 3.1 vector (AV:N/AC:H/PR:N/UI:N/S:C/C:L/I:L/A:N, 5.4) reflects that attacker-controlled precondition.
Arbitrary shortcode execution in the ProfilePress (formerly WP User Avatar) WordPress plugin through version 4.17.5 allows authenticated users with Subscriber-level access or higher to plant shortcode markup inside a comment, which is later rendered server-side by do_shortcode() when any visitor opens that user's front-end profile Comments tab. Because the plugin's ppress_neutralize_buffer_shortcodes() pass only strips pp-custom-html and bracket fragments containing HTML delimiters, shortcodes exposed by any other active plugin or theme remain intact and can be executed with attacker-chosen attributes, giving a low-privilege user a route to capability-gated or data-returning shortcodes and to raw HTML injection on a publicly viewable page. Rated CVSS 5.4 (AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N), the impact is bounded by which shortcodes a given site happens to have installed, and no public exploit code or confirmed active exploitation was identified at time of analysis.
Arbitrary shortcode execution in the Slim SEO WordPress plugin (all versions up to and including 4.11.1) allows an authenticated attacker holding a Subscriber-level account to store a shortcode payload in their own Biographical Info field, which the plugin subsequently passes to do_shortcode when rendering that author's description on the front end. Exploitation is gated by concrete prerequisites - an attacker-controlled low-privilege account (in practice, open user registration), a front-end request that resolves the author description without redirection (the ?author= query parameter while Slim SEO's auto_redirection setting is disabled), and the presence of other plugins whose shortcodes are worth invoking - and the resulting impact depends on which shortcodes other installed plugins expose, making this a moderate, authentication-gated code-injection issue rather than a top-priority emergency. No public exploit identified at time of analysis, and the vendor shipped a fix in 4.11.2 that strips shortcodes from the user bio.
An insecure direct object reference (IDOR/BOLA) in Tandoor Recipes prior to 2.6.10 lets an authenticated member of a shared Space read and destroy another member's pending bookmarklet import. By submitting a co-tenant's BookmarkletImport primary key to POST /api/recipe-from-source/, the attacker causes the endpoint to return the victim's stored source URL and captured page content and then unconditionally delete that pending record, yielding limited confidentiality loss and a low-impact availability/denial effect against the victim's import workflow. Exploitation requires a valid account in the same shared Space as the victim (CVSS 3.1 5.4, PR:L, no user interaction), exposing only multi-member deployments; single-user or non-shared Spaces are not affected, and no public exploit code was identified at time of analysis.
Missing FILE-privilege and secure_file_priv enforcement in MariaDB Server's optional CONNECT storage engine plugin lets an authenticated, low-privileged database user read or write files on the database host through the file-access UDFs json_file(), bson_file(), jfile_make(), bfile_make(), and jbin_file(), bypassing the access controls those functions were supposed to honor. Affected releases range from 10.6.1 up to (but excluding) the fixed builds 10.6.28, 10.11.19, 11.4.13, 11.8.9, 12.3.3, and 13.0.2; exploitation requires the CONNECT plugin to be installed and loaded (it is not part of a default MariaDB installation) plus a valid low-privileged account able to call the affected UDFs. The practical impact is bounded by the operating-system permissions of the database service account, so reachable files are limited to those the mysqld user can already read or write; no public exploit code has been identified at time of analysis, and the flaw is rated moderate (CVSS 3.1 5.4, AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N) rather than an emergency.
Server-side request forgery in SparkyFitness (self-hosted AI health and fitness tracker) before version 1.6.0 lets an authenticated user point the built-in custom AI service configuration at arbitrary internal or private network addresses, causing the server to issue outbound POST requests carrying an attacker-controlled bearer token and to leak the AI system prompt. Exploitation requires a valid account and the ability to save a custom AI provider URL (custom, ollama, or openai_compatible) and then trigger an AI flow such as chat or the label/food-photo scan, matching the assessed vector CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N (5.4). No public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis; impact is bounded to internal-service probing, forged outbound POSTs, and prompt disclosure rather than code execution. The issue is fixed in 1.6.0, which enforces an outbound URL policy that rejects private-network AI destinations with HTTP 403 by default.
Authenticated attackers holding only subscriber-level access can execute arbitrary WordPress shortcodes on sites running the WPEverest User Registration & Membership plugin in all versions up to and including 5.2.8, because an action passes an unvalidated value straight into do_shortcode without checking it first. The flaw is rated CVSS 5.4 (AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N) and there is no public exploit identified at time of analysis - it is not confirmed actively exploited (CISA KEV). Impact is bounded, but the practical risk is elevated on the many sites that leave open user registration enabled, since an attacker can simply self-provision the required subscriber account and then trigger the vulnerable shortcode path; sites that disable open registration meaningfully reduce exposure.
IBM i releases 7.3, 7.4, 7.5 and 7.6 can allow a remote authenticated attacker to slip past the ChrootDirectory confinement that is supposed to isolate SSH/SFTP users, exposing files and data outside the directory they were meant to be restricted to. The weakness is entirely dependent on configuration: it only affects systems that explicitly set ChrootDirectory in sshd_config and rely on it as an isolation boundary, while default deployments without chroot confinement are not impacted. No public exploit code has been identified at time of analysis, and because a valid restricted-account credential is mandatory and the assessed impact is bounded to low confidentiality and integrity loss with no availability effect, this rates as a moderate, operator-dependent issue rather than an emergency.
Account takeover in Apache Answer through 2.0.2 is possible because email verification codes issued for one workflow are accepted by other, more sensitive workflows - the code's original purpose is not enforced at redemption time. An attacker who obtains or replays a currently valid code from a lower-sensitivity email flow can present it to a sensitive flow tied to a target account, yielding high confidentiality and integrity impact (assessed CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:N, rated moderate by the project). Exploitation is constrained rather than trivial: the deployment must be using Apache Answer's email-code-based workflows and the attacker must hold a valid, unexpired code targeting a specific account workflow; no public exploit code or confirmed active exploitation was identified at time of analysis.
In applications that use the Microsoft Identity Web library for authentication and authorization, an attacker who already holds valid low-privilege credentials can tamper with authentication data the library incorrectly assumes is immutable, bypassing authorization checks to read data beyond their granted permissions (CVSS 5.3, C:H/I:N/A:N). Exploitation is not unauthenticated and is constrained by high attack complexity (AC:H), meaning specific conditions, state, or timing must be satisfied rather than a simple repeatable request. At the time of analysis there is no public exploit identified and no confirmed active exploitation, but a vendor patch is available (RL:O), so patched deployments are not affected.
Stored cross-site scripting in the Webhook URL form field of Crafty Controller allows an authenticated user with access to that configuration field to persist malicious input that executes in the browser of any other user who later views the affected page. The vulnerability is assessed at CVSS 5.4 (CVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:C/C:L/I:L/A:N), meaning a valid low-privileged account is required (PR:L) and a separate victim must load the rendering view (UI:R); the attacker cannot trigger impact on their own session, and no public exploit code was identified at time of analysis. Exploitation is most damaging when a higher-privileged operator such as an administrator views the poisoned webhook view, but this is a genuinely moderate, not high-priority, issue and no anonymous or pre-authentication path exists.
Authorization bypass in SmartCrawl (smartcrawl-seo) through version 3.16.4 allows an authenticated low-privilege user to manipulate a user-controlled key and access or modify data they should not be able to access, due to CWE-639 insecure direct object reference and incorrect access control configuration. This is a moderate-severity issue (CVSS 5.4, AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N) whose real-world risk is constrained by the requirement for a valid low-privilege account; no public exploit identified at time of analysis, and no KEV or EPSS signal was provided. Sites running affected SmartCrawl versions should monitor the Patchstack advisory and update beyond 3.16.4 once a fixed release is confirmed.
Remote denial of service in libIEC61850 1.6.1 allows an unauthenticated network attacker to crash an MMS server that has the file service and ObtainFile operation enabled. A successful ObtainFile frees the task's file handle but leaves the stale non-NULL pointer in a task reset to the NOT_USED state; a subsequent ObtainFile that reuses that task and fails destination-file creation then closes the already-freed handle a second time (CWE-415 double free), aborting the server process. Exploitation requires a specific success-then-failure request sequence and reachability to the exposed MMS endpoint; no public exploit code has been identified at time of analysis, and impact is limited to availability (CVSS 5.3, AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L) with no data exposure. The flaw is confirmed fixed in version 1.6.2.
Sensitive information exposure in HUSKY - Products Filter for WooCommerce Professional through 1.4.5 allows unauthenticated attackers to extract the external MySQL host, database name, username, and plaintext password from the inline woof_tm_data script object on public frontend filter pages when the Statistics extension, saved external MySQL credentials, and Turbo Mode are all enabled. This is a moderate, conditional risk: default or internal-database configurations are not affected, and no public exploit was identified at time of analysis, but any site that has adopted those documented settings should treat the credentials as potentially exposed.
Authenticated file disclosure in Helical Insight through 7.0.0 lets a user with viewer-level privileges read files outside the solution directory through the path parameter of /getExternalResource, bypassing per-resource access control by supplying parent-directory segments. Exploitation requires a valid low-privilege session and is constrained to files carrying an allowed extension (js, json, html), so confidentiality impact is limited and there is no integrity or availability effect - matching the assessed CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N and the assigned score of 5.3. No public exploit code was identified at time of analysis, the CVE is not listed in CISA KEV, and no EPSS probability was supplied; this is a medium-severity information-disclosure issue rather than an urgent priority.
Unbounded decoding of PMAPPROC_DUMP replies in libtirpc lets a malicious or already-compromised RPC portmapper exhaust the memory and CPU of a client application that calls pmap_getmaps(), crashing or stalling the client process and causing a denial of service. The exposure is conditional and client-side: a remote attacker cannot initiate the exchange, so the victim must direct libtirpc-based tooling (NFS/RPC mount helpers, discovery or inventory scanners) at the attacker-controlled rpcbind, which is why the assessed vector is AV:N/AC:H/PR:N/UI:R with availability-only impact and no confidentiality, integrity, or code-execution consequences. Affected packages are those shipped in Red Hat Enterprise Linux 6 through 10, Red Hat Hardened Images, and Red Hat OpenShift Container Platform 4. At CVSS 5.3 this is a moderate-severity, low-priority issue in practice, and no public exploit code and no confirmed active exploitation (no CISA KEV listing) were identified at time of analysis.
Missing authorization in the Piranha CMS manager API (Piranha.Manager LanguageApiController) lets any authenticated manager user list, create, edit, and delete site languages even when their account lacks the granular Language, LanguageAdd, LanguageEdit, and LanguageDelete permissions. Affected deployments are Piranha CMS through version 12.2; the attacker must already hold a valid manager account and simply call the manager/api/language Get, Save, and Delete actions directly. Impact is confined to language configuration and the localized content that depends on it - deleted languages cause affected sites to fall back to the default language instance-wide, with no confidentiality exposure; no CVE is confirmed as actively exploited (CISA KEV) and no public exploit identified at time of analysis, though a public technical writeup and vendor advisory are available.
Authenticated low-privileged users in Exrick XBoot through 3.3.4 can bypass role-based data-scope restrictions and read user records from any department by sending GET /xboot/user/getByDepartmentId/{departmentId} with arbitrary department IDs. The endpoint fails to enforce the data-scope filter, exposing usernames, real names, mobile numbers, emails, and account status (CWE-639 IDOR). The flaw has a moderate CVSS 3.1 score of 5.3 (medium) and requires only a valid session (PR:L, no user interaction); no public exploit code or active exploitation has been identified at time of analysis.
Insecure direct object reference (IDOR) in FastBee through 2.1.0 allows authenticated low-privilege users-including self-registered accounts-to read, overwrite, reassign device associations of, or delete other users' device groups by supplying sequential group IDs, because GroupController queries omit owner/tenant checks. This is a moderate-severity authorization flaw (CVSS 5.3) confined to device-group objects, with no code execution or full account takeover. No public exploit code has been identified at time of analysis, and the issue is not listed in CISA KEV.
Authenticated, low-privileged users of ClickVisual through 1.1.2 can reach ClickHouse instances they were never granted by substituting or incrementing the instance-id parameter in the /api/v2/reports handlers (ReportSourceDatabases, ReportSourceTables, ReportTableColumns, and WhereCheckRun). The handlers in api/internal/api/apiv2/report/report.go fail to confirm that the caller holds a grant on the referenced ClickHouse instance (CWE-862), so any logged-in account can enumerate database and table schemas and infer row counts of restricted tables; the issue is rated 5.3 (CVSS 4.0) and assessed as CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N. Exploitation requires a valid low-privileged account (it is not reachable anonymously), needs no special configuration or user interaction, and is limited to metadata disclosure rather than table contents - no public exploit identified at time of analysis, though a public technical writeup of the flaw is referenced by the CVE.
Cross-tenant data exposure in IoTSharp through 3.6.1 lets any authenticated user read other tenants' customer contact details, device inventories, and device attribute telemetry because the Customers and Devices API handlers fail to scope queries to the caller's tenant (CWE-639, CVSS 5.3). Authentication is trivial to obtain: the anonymous AccountController.Create endpoint allows self-registration, so an attacker only needs a valid GUID belonging to another tenant, and although those GUIDs are random 128-bit values that are not sequentially enumerable, exploitation is realistically targeted rather than mass-enumerable. Impact is read-only and limited to confidentiality - no integrity or availability effect - and the deployment must be genuinely multi-tenant to be exploitable; no public exploit identified at time of analysis, and there is no CISA KEV confirmation of active exploitation.
Missing department-level authorization in PlayEdu through 2.3 lets an authenticated student read course details and download attachment files that belong to departments the student is not entitled to see, because the frontend CourseController skips the department visibility check on those endpoints. The flaw is an IDOR-style (CWE-639) access-control gap: course and attachment identifiers are sequential and therefore enumerable, so a low-privileged account with valid credentials can walk the ID space and harvest course structure and files from other departments. Impact is confidentiality-only - no integrity or availability effect and no escalation to administrative privileges - and the vendor CVSS 4.0 base score of 5.3 (independent assessment CVSS 3.1: AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N) reflects a medium-severity information-disclosure risk. Exploitation status: no public exploit identified at time of analysis; a public third-party technical write-up of the flaw has been published, and no CISA KEV listing has been associated with this CVE.
Tbed instances through the 20240319 snapshot let any authenticated user rename another user's images by submitting the victim's imgname to the /admin/setImgFileName endpoint, a missing-authorization (CWE-639) flaw that undermines integrity but not confidentiality or availability. Because imgname values are exposed in public image URLs and album listings, the attacker needs no privileged knowledge to target a victim, and the write allows attacker-chosen display text to be attributed to that victim's images (impersonation/defacement). The assessed CVSS 3.1 vector is AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:L/A:N (scored 5.3 in the source data), requiring a valid low-privilege account but no victim interaction; no public exploit identified at time of analysis.
Authenticated Contributor-level users on a WordPress multisite network running 10up Distributor before 2.3.2 can create draft or published posts on network sites they do not have access to by invoking the dt-push AJAX action with a chosen connection ID and postStatus. The flaw is a moderate-severity missing authorization condition in ajax_push() (CWE-862) and is limited to multisite installations; single-site WordPress installs are not exploitable, and super admins are out of scope. Vendor-released patch: 2.3.2; no public exploit identified at time of analysis.
Unauthenticated content injection in the WordPress NotificationX plugin (WPDevTeam) - versions up to and including 3.3.3 - allows remote attackers to write forged social-proof entries (name, title, message, email, theme) into any enabled non-popup NotificationX campaign by POSTing an arbitrary nx_id to the /notificationx/v1/popup-submit REST route. That route is registered with permission_callback => '__return_true', so no authentication, user interaction, or non-default plugin setting is needed, although exploitation does depend on at least one non-popup campaign (WP Comments, WooCommerce/EDD/Give sales, reviews, etc.) already existing and being enabled on the target site. The impact is confined to the integrity of displayed notification content (CVSS 5.3 Medium - no confidentiality or availability loss), no public exploit code has been identified at time of analysis, and the issue is not flagged as actively exploited.
Missing authorization in Allegro Ralph through 20260902.1 lets any authenticated staff-level user read data-center asset details and manipulate racks through the API. Because RalphPermission skips checks for views that lack a queryset, the DCAssetsView and SRRacksAPIView endpoints at /api/rack/<id> and /api/server_room/<id> are reachable without the rack model permissions they should require, exposing asset hostnames, serials and management IPs and allowing rack creation, modification or movement. No public exploit code has been identified at time of analysis, and the flaw is rated medium severity (CVSS 4.0 base 5.3; independent CVSS 3.1 base 5.3) with risk gated by the need for a valid low-privilege staff credential.
AIHelms 0.1.25 and earlier fails to enforce efficiency:read and efficiency:write authorization in its analytics API, allowing any authenticated low-privileged user to read organization-wide AI spend data and to modify suggestion statuses or generate reports. The exposed routes include /cost/detail, /top-users and /budget in apps/api/v1/efficiency.py, and exploitation requires only a valid low-privilege account - no elevated permissions and no user interaction. Independent assessment rates this a genuine but moderate-priority issue (assessed CVSS 3.1 5.3 with PR:L, C:L/I:L/A:N); no public exploit code or confirmed active exploitation was identified at time of analysis.
Missing authorization in DooTask through 1.9.66 allows authenticated users to read and inject messages into OKR comment dialogs by supplying the dialog_id to msg__list, msg__one, or msg__send* endpoints. The flaw, CWE-639 (IDOR), skips the membership check in WebSocketDialog::checkDialog() for dialogs with group_type 'okr'. Exploitation requires a valid account and knowledge of the target dialog_id; impact is limited to low confidentiality and integrity loss (CVSS 5.3). No public exploit code or active exploitation has been identified.
Stored cross-site scripting in Backdrop CMS Core 1.35.x before 1.35.2 and 1.34.x before 1.34.6 allows an authenticated user holding SVG upload permission to embed dangerous tags or scripting inside an uploaded SVG image, which will execute in a victim's browser only if that SVG is opened directly by its raw file URL. Backdrop's design of embedding all uploaded SVGs inside <img> tags blocks that script during normal page rendering, and reaching the payload requires both the upload privilege (PR:L per the assessed vector) and a victim who navigates to the file itself, so this is a moderate-severity issue (CVSS 5.3, CWE-79) with no public exploit identified at time of analysis and no confirmed active exploitation (CISA KEV).
Missing authorization in Ekko Studio through 0.7.36 allows any authenticated user who is not a super_admin to list and clear server-wide login lockouts through the /api/auth/locked-ips endpoint, with a DELETE request lacking an ip parameter invoking the unlockAll() routine. The practical consequence is that the application's login rate limiter can be reset at will, removing the primary obstacle to sustained password guessing against super_admin accounts; the flaw requires a valid low-privileged account (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N, medium severity), so it is most relevant to insider or post-compromise scenarios rather than fully unauthenticated external attack, and it yields no direct account takeover on its own. No public exploit identified at time of analysis (no CISA KEV listing and no exploit code referenced); vendor-supplied fix data is present.
Authenticated users of abp-next-admin 10.2.0 through 10.6.0 who hold the Chat permission can read, and inject messages into, other users' AI chat conversations by supplying a victim's ConversationId GUID to ChatAppService.GetListAsync or SendMessageAsync in the LINGYUN.Abp.AIManagement module. The flaw is a moderate-severity authorization error (CWE-639, IDOR) rather than an unauthenticated remote compromise: exploitation requires a valid account holding the Chat permission (PR:L per the assessed vector) plus knowledge of the target's GUID, which is not readily guessable, so opportunistic abuse is limited. There is confirmed actively exploited status absent (CISA KEV), no public exploit identified at time of analysis, and exploitation is constrained by the out-of-band acquisition of the ConversationId; the advisory and vendor source code are publicly referenced, but that is technical documentation, not weaponized exploit code.
Missing authorization in B3log Solo through 4.4.0 lets an authenticated author account read and act on articles belonging to other authors, because the ArticleConsole handler never enforces per-author ownership checks. The assessed impact is limited confidentiality and integrity loss (CVSS 3.1 5.3, PR:L), and exploitation requires a valid author-level account on a deployment with more than one author, so single-author blogs have nothing to cross-access; the console articles-status endpoint can be queried with the k parameter to probe other authors' drafts, and push2rhy can be used to publish others' articles into the Rhythm community. No public exploit code or confirmed active exploitation was identified at time of analysis, and no vendor-released patched version has been identified.
Missing object-level authorization in MrDoc through 1.1.1 lets any authenticated user publish other users' private drafts, reparent their documents, and inject documents into projects they do not own by supplying arbitrary document or project identifiers to the ImportLocalDoc API. The flaw is a classic insecure direct object reference (CWE-639) assessed at 5.3 (CVSS:4.0 AV:N/AC:L/PR:L/UI:N, low confidentiality and integrity impact) and is confined to the application's own authorization boundary: it cannot reach the host, and it requires a multi-user deployment in which a victim's document or project actually exists. No public exploit code was identified at time of analysis and the CVE is not listed in CISA KEV, though a public technical writeup documents the endpoint's behavior.
Sensitive information exposure in the WordPress plugin Lightweight Cookie Notice (daext) in all versions up to and including 1.19 leaks the W3TC_DYNAMIC_SECURITY constant, which the plugin hardcodes to a publicly known value; a remote, unauthenticated attacker can therefore learn the W3 Total Cache mfunc security key without any credentials or user interaction. On sites that also run W3 Total Cache with its mfunc dynamic-content feature active, knowledge of that key may allow an attacker to forge signed mfunc directives and inject arbitrary PHP code, but the confidentiality leak alone scores only 5.3 (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N) and escalation to code execution is strictly conditional on the co-presence and configuration of that second plugin. No public exploit code has been identified at time of analysis and there is no indication of confirmed active exploitation (no CISA KEV listing).
Missing authorization in Arkon through 0.9.2 lets any authenticated employee who holds only the wiki:read:own_dept permission read another department's wiki metadata - page slugs, titles, node types and link edges up to depth three - by supplying a foreign department slug to GET /api/wiki/graph. Root cause is a CWE-639 user-controlled key flaw: the route passes the caller-supplied slug straight into get_neighborhood() without verifying that the requested department belongs to the requester, so an ordinary employee-level account is sufficient and no elevated role is needed. Impact is deliberately narrow (CVSS 5.3, and our own assessment of CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N): read-only metadata disclosure only, with no page body content, no data modification and no availability impact; no public exploit identified at time of analysis, and exploitation merely requires knowing or guessing a target department slug.
An insecure direct object reference in Fuint through commit 57ec018 lets authenticated members retrieve other members' booking records by supplying arbitrary bookId values to /clientApi/book/myBookDetail. Because bookId is an auto-increment integer, an attacker with a valid member account can enumerate records and obtain contact names, mobile numbers, in-store verification codes, service dates, stores, and remarks. No public exploit identified at time of analysis and no confirmed active exploitation; assessed severity is moderate with low confidentiality impact and no integrity or availability impact.
Exposed debug/diagnostic interfaces in SFM Software Technologies Inc.'s Costifier application let unauthenticated remote attackers reach non-production code paths in production deployments, resulting in limited information disclosure. The flaw carries a medium-severity CVSS 3.1 base score of 5.3 (AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N), meaning no credentials, no user interaction, and no special conditions are needed once the debug interface is network-reachable; integrity and availability are unaffected. All Costifier versions released through 2026-10-16 are in scope, the vendor was contacted but did not respond, and no public exploit identified at time of analysis.
Distributed transaction manipulation in IBM i 7.3, 7.4, 7.5, and 7.6 stems from improper validation of attacker-supplied transaction identifiers, allowing a remote, unauthenticated attacker to alter the state of distributed (two-phase commit / XA-style) transactions that the system participates in. The impact is limited to integrity (CVSS 3.1 base score 3.7, AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N) with no confidentiality or availability effect, and high attack complexity means the attacker must know or predict valid transaction identifiers rather than send arbitrary input. No public exploit code has been identified at time of analysis and there is no confirmed active exploitation; a vendor patch is available, so patched systems are not exploitable.