Missing server-side enforcement of the write/submission password in TDuck survey form through 6.0 allows remote unauthenticated attackers to post form entries directly to the public submission API, because the password gate is implemented only in front-end JavaScript. An attacker only needs the form key that is exposed in any public share link; no account, no user interaction, and no special configuration are required. Publicly available exploit code exists (a Python PoC targeting the submission endpoint), but there is no indication of confirmed active exploitation (not in CISA KEV), and the impact is bounded to unauthorized submission injection - existing responses cannot be read and availability is unaffected.
Denial of service in containerd versions prior to 1.7.36, 2.0.13, 2.2.9, 2.3.6, and 2.4.1 allows an unauthenticated attacker who controls or poisons an image source to force excessive CPU and memory consumption during PullImage, causing long ContainerCreating stalls and potential node instability at larger graph sizes. The vulnerability is triggered before container start and requires no privileges on containerd itself, but is limited to environments that pull untrusted images; clusters pulling only trusted first-party images are not exposed. No public exploit identified at time of analysis, and impact is availability-only with no confidentiality or integrity effect.
Unauthenticated credential disclosure in X-SpringBoot (yzcheng90) through version 6.0 allows remote attackers to read tenant appKey and appSecret values from the GET /application/manager/select endpoint, which performs no authentication check and applies no field filtering to the response. With those credentials an attacker can drive the affected tenant's configured SMS provider to send arbitrary messages, enabling SMS bombing, cost abuse, and impersonation of the legitimate application. The issue is trivially exploitable - the assessed vector is CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N (vendor score 6.9 under CVSS 4.0) - and publicly available exploit code exists in a GitHub PoC repository; there is no indication of confirmed active exploitation (not in CISA KEV) and no vendor-released patch was identified at time of analysis. Direct impact is bounded: the disclosure affects SMS credentials and message-sending capability rather than full system compromise.
Reflected cross-site scripting in the Link Library WordPress plugin (versions before 7.9.6) lets remote, unauthenticated attackers inject script via parameters that are reflected, unescaped, into the link addresses generated on the plugin's front-end directory pages. Exploitation requires the victim to open an attacker-crafted URL (CVSS UI:R), so delivery depends on phishing or social engineering, and the impact is substantially greater when the tricked user is a logged-in administrator because the injected script runs with that session. Publicly available exploit code exists, but this CVE is not in CISA KEV and EPSS is low at 0.15% (4th percentile), so broad real-world exploitation is not currently suggested; a vendor-released patch is available in 7.9.6.
TDuck Survey Form 6.0 fails open in its form-ownership authorization check, letting an authenticated user read submission records that belong to forms which have been permanently deleted. An attacker who knows or can guess a dataId can call GET /user/form/data/details and retrieve the orphaned submission row, including personal information, because FormAuthUtils.hasPermission returns a permissive result when the parent form no longer exists; for live forms the ownership check behaves normally. Exploit code is publicly available, no CISA KEV entry indicates active exploitation, and the practical risk is bounded by the need for valid credentials, a known or enumerable dataId, and the strict precondition that the parent form has already been permanently deleted - for these reasons the real-world impact sits below the headline 6.0 CVSS 4.0 score.
Unauthenticated access to the Swagger/OpenAPI api-docs endpoint in zhistaredu StarTraining through 3.8.1 allows remote attackers to retrieve API documentation and structural details when the endpoint is network-reachable and SecurityConfig.java lacks an authentication filter. Public exploit code exists, but this is an information-disclosure issue (assessed CVSS 3.1: AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N), not confirmed active exploitation; the vendor did not respond to the disclosure. The practical risk depends on exposing the api-docs route, and impact is limited to confidentiality of API structure.
Unauthenticated remote access to the Alibaba Druid monitoring console bundled with zhistaredu StarTraining 3.8.1 and earlier allows anyone who can reach the /druid/index.html endpoint to view monitoring data and perform console reset operations without credentials. No user interaction or privileges are required (CVSS vector PR:N/UI:N), though real-world exposure depends on the Druid StatViewServlet being enabled and network-reachable without an authentication filter - a non-default but commonly insecure configuration. Publicly available exploit code exists (a GitHub advisory with reproduction details), but there is no evidence of confirmed active exploitation, and the vendor was contacted and did not respond.
Code injection in the Zhonglun CloudPos desktop client (versions up to and including 3.0.1.76) allows remote attackers to inject and execute code through the 'url' argument of the OpenLocalBrowser function in the CefSharp-based JSBridge component (ZlPos/ZlPos/Bizlogic/JSBridge.cs), with the vendor's own CVE record pointing to RCE as the practical outcome. Publicly available exploit code exists - a proof-of-concept document titled 'CloudPos.CefSharp.RCE' is linked from the disclosure - so the attack is both demonstrable and reproducible. The vendor was contacted but did not respond, so no vendor-released patch has been identified; exploitation requires the affected desktop client to process an attacker-influenced url (the independent assessment rates UI:R, implying some operator interaction or attacker-controlled content delivery is likely despite the CVSS 4.0 vector's UI:N claim). The CPE entry uses a wildcard version range, so all released builds up to 3.0.1.76 should be treated as affected.
SQL injection in the faculty authentication handler (loginlinkfaculty.php) of mathurvishal's CloudClassroom-PHP-Project lets unauthenticated remote attackers manipulate the fid or pass request arguments to alter the backend SQL query, potentially bypassing faculty login and reading or modifying application data. Exploitation needs only network reachability to the exposed login endpoint - the CVSS 4.0 vector is AV:N/AC:L/AT:N/PR:N/UI:N with Low confidentiality, integrity, and availability impact (base 6.9), so no credentials, user interaction, or non-default configuration are required. Publicly available exploit code exists (a GitHub proof-of-concept demonstrating authentication bypass), but the practical risk is tempered by deployment scarcity: the vendor uses rolling releases, published no version details, responded to neither disclosure, and no vendor-released patch is currently identified.
StarTraining 3.8.1 and earlier embeds a hard-coded JWT signing secret in application.yml, allowing an unauthenticated remote attacker who knows that static value to forge valid authentication tokens carrying attacker-chosen user_id and company_id claims, effectively impersonating arbitrary users without any credentials. Publicly available exploit code exists (a GitHub advisory demonstrates forging tokens from the shipped secret), but the flaw is not in CISA KEV and no confirmed active exploitation has been reported. The vendor was contacted about the disclosure and did not respond, so no vendor-released patch is identified; installations that have rotated the JWT signing secret away from the shipped default are not exploitable through the disclosed value.
SQL injection in mathurvishal's CloudClassroom-PHP-Project (updatequery.php) lets remote attackers manipulate the gid parameter to inject arbitrary SQL into the application's database queries, rated CVSS 4.0 AV:N/AC:L/AT:N/PR:N/UI:N with low confidentiality, integrity, and availability impact. Publicly available exploit code exists via a GitHub advisory (proof-of-concept maturity per the E:P vector), but there is no CISA KEV listing and no evidence of active exploitation; CISA's SSVC assessment is Exploitation: poc, Automatable: yes, Technical Impact: partial. The vendor was contacted early and did not respond, and the project uses rolling releases, so no fixed version can be identified; combined with the product's near-zero deployment footprint, real-world risk is LOW.
Authorization bypass in ningzichun's student-management-system (all versions up to commit 98760f5711cf6dc8b4adca53a9e207ca49b02ebf) allows a remote attacker to tamper with log records by manipulating the sid, addtime, type, reason, detail and logdate arguments passed to user/editLog.php, circumventing the intended ownership/authorization check on those records. Publicly available exploit code exists and the flaw was disclosed through a public GitHub issue report that the maintainer has not responded to, so exposed deployments should be considered at risk. The impact is integrity-only log tampering with no confidentiality or availability loss, and because the affected code is a rare, unmaintained hobby project with no patched release, the realistic risk is low despite a moderate 6.9 CVSS 4.0 base score; authentication requirements are not firmly confirmed from available data, as VulDB rates the vector PR:N while independent analysis assesses PR:L.
Authorization bypass in ningzichun student-management-system (versions up to commit 98760f5711cf6dc8b4adca53a9e207ca49b02ebf) allows remote attackers to read other students' records by manipulating the 'sid' parameter of admin/fun/getStudent.php. Publicly available exploit code exists, and the maintainer has not responded to the report, so no vendor-released patch is identified. The assessed CVSS 3.1 vector is AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N, limiting impact to confidentiality of individual records, but the PR:N rating assumes the admin/ path is reachable without upstream authentication.
SQL injection in the updatefaculty.php script of mathurvishal's CloudClassroom-PHP-Project allows remote attackers to inject arbitrary SQL through the unsanitized fid parameter, potentially reading or altering faculty records stored in the underlying database. Publicly available exploit code exists, published as a GitHub advisory with a SQLi proof-of-concept, and the vendor was contacted but never responded, so no fixed version is available; the CPE record (cpe:2.3:a:mathurvishal:cloudclassroom-php-project:*) carries a wildcard version because the project uses a rolling-release model. Practical risk is bounded by the product's negligible production footprint as an obscure educational PHP project rather than by attack difficulty, making this a real but low-priority issue.
Authenticated non-admin users can add, modify, or delete global form themes and theme categories in TDuck Survey Form through 6.0 because the FormThemeController write endpoints enforce neither an administrative role nor ownership of the affected theme resources. The flaw is a missing-authorization (CWE-862) issue reachable by any holder of a valid low-privilege account (CVSS 4.0 vector PR:L, AV:N, AC:L, UI:N) with no confidentiality impact - only integrity and availability effects on themes shared across other users' forms. Publicly available exploit code exists (a proof-of-concept script targeting the theme write path), but there is no confirmed active exploitation in the wild; the NVD score of 5.3 is credible and consistent with this limited, credential-gated impact.
Authenticated users who know a form's key can flood TDuck Survey Form through 6.0 with unlimited submissions via POST /user/form/data/create, because that endpoint never enforces the form's publish status, time window, quota, or per-IP limits - letting an attacker falsify or inflate collected survey results. The defect requires only a low-privileged TDuck account (CVSS 3.1 PR:L) plus the target form key, is remotely exploitable over HTTP with no user interaction, and is strictly an integrity issue: no confidential data can be read and no denial of service results. Publicly available exploit code exists (a bypass PoC script for the data-create endpoint), but there is no confirmed active exploitation and no CISA KEV listing; independent assessment rates this a genuine but low-severity authorization defect.
Rate-limit bypass in the wpForo Forum WordPress plugin (3.0.0 up to but not including 3.1.6) allows unauthenticated remote attackers to drain a site owner's paid, metered AI credits by spoofing the client-supplied IP header the plugin trusts when keying its per-visitor throttle on AI requests. Exploitation requires the plugin's paid/metered AI feature to be enabled and funded, and the deployment must accept forwarded-IP headers from untrusted clients; there is no code execution, data disclosure, or privilege gain, with impact limited to financial/resource exhaustion (CVSS 5.3, A:L), and publicly available exploit code exists, though there is no confirmed active exploitation (not in CISA KEV). Sites that have not enabled the AI feature or that normalize forwarded-IP headers at a trusted proxy are not meaningfully exposed.
Unauthenticated customer record tampering in the Bookly WordPress plugin before 28.3 lets remote attackers overwrite a known customer's stored personal details - name, email and address - because the plugin fails to verify the requester's identity before processing a customer-details update. Exploitation is trivial and needs no login or user interaction (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N, 5.3), but it is not blind: the attacker must already hold a valid customer's primary identifier, which confines practical abuse to targeted modification of a specific known record rather than mass, automated takeover of the whole customer base. Publicly available exploit code exists and WPScan has published a proof-of-concept advisory, though there is no indication of confirmed active exploitation; integrity impact is limited (I:L) with no confidentiality or availability loss, and this is a genuine but moderate-severity issue rather than a top-tier priority.
Blind server-side request forgery in OpenMetadata through 2.0.2 allows an authenticated user holding permission to create or update an EventSubscription to point a WEBHOOK destination at internal hosts, loopback addresses, or the cloud metadata endpoint 169.254.169.254, because URLValidator.validateURL neither resolves DNS hostnames nor rejects internal/private addresses. Only the resulting HTTP status code is returned to the caller, so the leak is blind internal probing rather than response-body exfiltration or code execution. Publicly available exploit code exists (a proof-of-concept gist is linked from the vendor advisory) and the assessed CVSS 4.0 base score is 5.1; no confirmed active exploitation (CISA KEV) is referenced in the available intelligence.
Authenticated MasterStudy LMS members can bypass membership-plan restrictions to enroll in paid courses outside their plan and exceed the plan's course-count limits in plugin versions from 1.9 before 3.7.50. The flaw is a server-side authorization failure in the enrollment flow: the plugin does not verify that the requested course is covered by the member's plan or that the submitted plan identifier actually belongs to that member. Publicly available exploit code exists, per the WPScan advisory, and the issue is rated low-to-moderate; the CVSS vector shows network access with a member-level account (PR:L) and low confidentiality/integrity impact, with no availability impact.
Link Library, a WordPress plugin, writes a generated image to a destination folder that is taken directly from user input without sanitization, so an authenticated user with the Contributor role or higher can create arbitrary directories and write or overwrite image files anywhere the web server process has write permission - including paths outside the site's document root (CVE-2026-78394, CWE-22 path traversal, affecting versions before 7.9.6). Because the written filename is always numeric with a fixed image extension, the primitive cannot be used to drop executable PHP, and the assessed impact is limited to low-integrity file write/overwrite (CVSS 4.1, AV:N/AC:L/PR:L/UI:N/S:C/C:N/I:L/A:N); this is confirmed not to be an actively exploited issue (no CISA KEV entry), publicly available exploit code does exist, and EPSS is negligible at 0.16% (4th percentile). The vendor has released a fix in 7.9.6, which is the straightforward and recommended remedy.
Server-side request forgery in the Link Library WordPress plugin versions 7.8.8 through 7.9.5 allows unauthenticated visitors to make the site issue outbound requests to hosts on its internal network and to observe from the response whether an internal service answered. The flaw is an incomplete fix for CVE-2025-68600 and only manifests when the plugin's protected 'safe request' is rejected and it falls back to an unprotected fetch, so sites must have published the plugin's public link submission form with reciprocal-link validation enabled to be exposed. Publicly available exploit code exists, but EPSS is only 0.13% (2nd percentile) and the independent assessment rates this as a genuinely low-priority, information-only issue with a non-default configuration prerequisite.
Remote manipulation of Open Web Analytics through 1.8.1 queue bookkeeping is possible through the Remote Event Queue endpoint when event forwarding is enabled, allowing an unauthenticated network attacker to collide with queue-item primary keys, set far-future delivery timestamps, or preload retry counters. The broader CWE-502 object-injection primitive is not reachable solely by an unauthenticated HTTP caller because the vendor patch confirms HTTP input cannot name a class for unserialize; full deserialization exploitation requires an attacker who can already write to the owa_queue_item table or the queue log file, such as via SQL injection, stolen database credentials, or a restored/tampered table. No public exploit identified at time of analysis; vendor-released patch is version 1.8.2.
Denial of service in RabbitMQ's management HTTP API lets an authenticated administrator crash the broker by submitting a user (or importing definitions) whose tags list contains roughly one million unique strings; each string is converted to a permanent Erlang atom until the bounded atom table is exhausted and the BEAM node dies. Affected versions are RabbitMQ 4.0.0 through 4.0.21, 4.1.x before 4.1.14, and 4.2.x before 4.2.7, and the attack requires the management plugin to be enabled and reachable plus full administrator privileges, so there is no unauthenticated or low-privilege path. No public exploit code identified at time of analysis, and vendor-released patches are available in 4.0.22, 4.1.14 and 4.2.7; this is a genuine but low-priority availability issue rather than a confidentiality or integrity risk.
Unauthenticated information disclosure in Flame (pawelmalak's self-hosted homelab dashboard) through 2.4.0 allows any client that can reach the HTTP interface to retrieve the complete application configuration in a single request to GET /api/config, exposing the stored weather API key and internal operational settings. The endpoint is mounted without authentication middleware and the controller serializes the entire config object without redacting secret fields, so no credentials or special conditions are required - the CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:N/UI:N/VC:L/VI:N/A:N, base 6.9) and the independent assessment (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N) both confirm remote, unauthenticated exploitation against default deployments, though confidentiality impact is bounded and there is no integrity or availability impact. There is no public exploit identified at time of analysis and no confirmed active exploitation, but the disclosure requires only a plain HTTP GET, so any internet-exposed instance should be assumed to have had its configuration read.
Unauthenticated information disclosure in Wikimedia Foundation's WikiLambda Extension 1.47.0-alpha allows remote attackers to access a limited amount of sensitive data from affected MediaWiki installations. The issue is CWE-200 exposure of sensitive information, with a CVSS 4.0 base score of 6.9 (low confidentiality impact only and no integrity or availability impact); the independent assessment confirms network reachability with no authentication or user interaction. No public exploit identified at time of analysis, and the upstream remediated code is on the master branch via Gerrit change 1320186, tracked in Phabricator T430601.
Unlimited password guessing against the two-factor authentication login step in Zammad (all versions before 7.1.2) lets unauthenticated network attackers brute-force account passwords without triggering the platform's normal failed-login lockout or rate limiting, and the endpoint's response discloses whether each guessed password was correct before the second factor is ever evaluated. The concrete impact is credential-validity disclosure and unlimited offline-style online brute forcing rather than authentication bypass, since the second factor still gates actual login; risk is highest for accounts protected by weak or reused passwords. Rated CVSS 4.0 base 6.9 (AV:N/AC:L/AT:N/PR:N/UI:N, confidentiality-only), which is consistent with an independent CVSS 3.1 assessment of AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N; no public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis, and the vendor has released a fix in 7.1.2.
Reflected cross-site scripting in the MediaWiki Cargo extension (versions through 3.9.4) lets remote, unauthenticated attackers execute arbitrary script in the origin of an affected wiki by convincing a victim to open a crafted URL while logged in. The defect is a CWE-79 improper-neutralization issue that the vendor rates 6.9 (Medium) under CVSS 4.0, whereas the independent assessment applies CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N, explicitly requiring victim user interaction that the vendor vector omits; impact is confined to the victim's browser session within the wiki's origin rather than the server host. No public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis, and exposure is limited to the subset of wikis where this optional extension is installed and enabled.
Reflected cross-site scripting in the MediaWiki Cargo extension (versions through 3.9.4) allows an unauthenticated remote attacker to inject script that executes in a victim's browser within the wiki's origin, but only after the victim is induced to open a crafted URL. The vulnerable surface is limited to wikis that have separately installed and enabled the optional Cargo add-on, since it is not part of MediaWiki core, and no public exploit code has been identified at time of analysis. With an EPSS-style risk profile consistent with a modest-severity issue and impacts limited to session-scoped confidentiality and integrity, this is best handled as routine patching rather than an emergency response.
Reflected cross-site scripting in the MediaWiki Cargo extension, affecting all versions through 3.9.4, allows an unauthenticated attacker to inject script that executes in the browser of any wiki user who follows a crafted link. Per the assessed CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N), exploitation requires victim interaction - the target must open the malicious URL - and impact is limited to the victim's session on the affected wiki, with the scope change reflecting that the injected content can affect resources beyond the vulnerable component. No public exploit code has been identified and there is no CISA KEV entry: no public exploit identified at time of analysis. Exposure is confined to wikis that have installed and enabled Cargo, a non-core, separately deployed extension, and the Cargo query/reporting functionality must be reachable.
Stored cross-site scripting (XSS) in the MediaWiki Cargo extension through version 3.9.4 allows attackers to inject persistent malicious scripts into wiki pages that execute in the browsers of users viewing the content. The flaw requires the ability to save content that populates Cargo-stored data (typically an authenticated editor, though open wikis with anonymous editing may also be vulnerable) and user interaction from a victim. The Wikimedia Foundation reported the issue, but no public exploit code has been identified, and no vendor patch is yet available.
Information disclosure in the MediaWiki FlaggedRevs extension through 1.46.0 allows unauthenticated remote attackers to obtain limited sensitive information that the extension fails to remove before storage or transfer, per CWE-212. Only MediaWiki installations with the optional FlaggedRevs extension enabled are affected; wikis without it are not exposed. No public exploit identified at time of analysis, and SSVC records exploitation as none, making this a moderate-severity, low-to-medium operational priority.
Zammad 7.0.3 and 7.1.1 return stored LDAP and Exchange integration credentials in cleartext through the integration administration API, allowing an authenticated administrator to read secrets the UI is supposed to mask. Because the vulnerable endpoints are the import 'job_try' responses, a user holding full admin rights - and only such a user - can recover the LDAP bind password, the Exchange EWS password, or a stored OAuth access token, plus raw request parameters that leak into the persisted ImportJob payload. Per the supplied vector the issue requires high privileges (PR:H) and results in confidentiality impact only, with no code execution or data modification; there is no public exploit identified at time of analysis, and the vendor-released fix is Zammad 7.1.2.
Crafted URL injection in the Onion module of the AIL Framework lets unauthenticated content publishers push arbitrary non-onion addresses into the crawler's task queue, degrading the integrity of the framework's target selection. The module used a performance shortcut that accepted any URL of exactly 69 characters ending in '.onion' as a valid onion target, then sliced the domain directly out of the string without proper hostname parsing, so an attacker who can place content in a source the crawler ingests (a monitored paste, leak site, or web feed) can steer the crawler toward unintended network resources. No authentication or user interaction is needed and default configuration is sufficient; the assessed vector (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N) confirms impact limited to low integrity with no confidentiality or availability effect, and there is no public exploit identified at time of analysis.
A stack (or heap) out-of-bounds write in kitty terminal versions 0.40.0 up to but not including 0.49.0 allows any program that can emit escape sequences to the terminal to corrupt a fixed-size four-codepoint buffer while rendering an OSC 66 text-sizing sequence, crashing the kitty process and taking down every window, tab, and child process with it. The flaw is in screen_handle_multicell_command(), which appends each codepoint of a grapheme cluster with lc.chars[lc.count++] = ch without a capacity check in the natural-width branch of the text sizing protocol; a payload carrying a grapheme cluster longer than four codepoints writes one 32-bit value past the buffer for each additional codepoint. Exploitation is local and requires the user to view attacker-influenced terminal output (PR:N, UI:R); the impact is confined to availability - no confidentiality or integrity effect and no evidence of code execution. Vendor-released patch: 0.49.0; no public exploit identified at time of analysis.
Secure Folder 1.2, an Android vault app distributed as com.securefolder.securevault on Google Play, stores files selected for its password-protected vault as unencrypted files inside the Android shared-storage tree, so any co-resident application or file manager holding read access to that path can enumerate, copy, and open the vaulted files without authenticating to Secure Folder. The exposure is strictly local: the CVSS 4.0 vector requires local access with low privileges (AV:L, PR:L) and no user interaction, and impact is confidentiality-only, matching our independent CVSS 3.1 assessment (AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N), which characterizes real-world risk as low-to-moderate despite the 6.8 base score. There is no public exploit identified at time of analysis and no CISA KEV listing, so this is a design-level information disclosure rather than an actively attacked flaw. Any deployment of version 1.2 with files added to the vault is affected by default, since the plaintext storage behavior is not the result of an optional setting.
mTLS holder-of-key certificate binding is not enforced in Keycloak's Standard Token Exchange V2 flow, so an attacker who already holds valid confidential client credentials can obtain a standard, unrestricted token that escapes the certificate-binding protection the deployment believed was in place. Affected distributions include Red Hat Build of Keycloak and Red Hat Single Sign-On 7, and exploitation is conditional on three things being true at once: the attacker must already possess a stolen client secret or key (the CVSS vector rates this PR:L - a precondition the flaw itself does not grant), the deployment must have Standard Token Exchange V2 enabled and in active use, and the deployment must be relying on mTLS holder-of-key binding as a hardening control. No public exploit code has been identified at time of analysis and there is no CISA KEV listing; the assessed vector (AV:N/AC:H/PR:L/UI:N, C:H/I:H/A:N) reflects a genuine but conditional security-control bypass rather than a mass-exploitable flaw, since certificate-bound tokens are a non-default opt-in configuration.
Privilege escalation in the Fine-Grained Admin Permissions v2 (FGAP v2) feature of Keycloak lets an authenticated delegated administrator with role-assignment rights grant a composite role whose transitive contents are never inspected, silently conferring full realm administrative control. Exploitation requires that FGAP v2 be enabled for delegated administration, a limited-admin credential holding role-mapping permissions (CVSS vector PR:H), and the presence of a composite role that transitively contains realm-admin privileges within that administrator's assignable scope; the high attack complexity (AC:H) reflects these preconditions rather than any exploitation difficulty, and the resulting impact is a complete confidentiality, integrity, and availability compromise of the realm (CVSS 6.6). No public exploit code and no confirmed active exploitation have been identified at time of analysis, so this should be treated as a genuine but conditional authorization-logic flaw rather than a remotely exploitable takeover.
Authenticated path traversal in the Universal Connector Oracle Wallet upload component of IBM Guardium Data Protection 12.2 lets a low-privilege remote user write files to arbitrary filesystem locations on the appliance. The attacker must already hold a valid account with access to the Universal Connector upload function (PR:L) and must craft a traversal sequence that lands a file somewhere useful, which is why the vector carries AC:H despite the high confidentiality and integrity impact. Exploitation has not been observed: CISA SSVC records Exploitation as none and there is no public exploit identified at time of analysis, so this is a genuine but moderate-priority issue rather than a widespread emergency - though the 'total' technical impact rating means a successful hit against a misconfigured deployment could be severe.
An authenticated user with a valid CSRF token can read Django password hashes, request metadata, and related-user attributes from Horilla HR/CRM mail-preview functionality in versions 1.0.0 through pre-1.6.0 and prior to 2.0.0. The get_mail_preview handlers at /recruitment/get-mail-preview/ and /employee/get-employee-mail-preview render a user-supplied mail body as a Django template while the full request object sits in the template context, so template attribute traversal such as request.user.password or request.META exposes the underlying data; the Django template sandbox limits this to information disclosure rather than code execution. Exploitation requires only a low-privileged authenticated session with a CSRF token (no public exploit identified at time of analysis), and the confidentiality impact is high because the leaked password hashes can be attacked offline or used to identify accounts for follow-on compromise; vendor fixes ship in 1.6.0 and 2.0.0.
Server-side request forgery in the Image API (v2) of OpenStack Glance allows an authenticated tenant user (CVSS vector PR:L, network reachable) to make the glance-api service issue HTTP requests to attacker-chosen URLs by manipulating the 'locations' attribute of an image that is in the 'queued' state via a crafted PATCH request, potentially exposing internal-only services or cloud metadata (high confidentiality impact, limited integrity impact per the assessed vector). The vulnerability is strictly gated: it only functions when the non-default, deprecated 'show_multiple_locations' option is explicitly enabled in glance-api.conf, so standard OpenStack deployments that leave this setting off are not exploitable, and the attacker must hold valid tenant credentials and network access to the Glance API endpoint. Scoring is 8.1 with a low EPSS of 0.31% (21st percentile); the issue is not confirmed actively exploited (CISA KEV) and no public exploit identified at time of analysis, although a GitHub repository referencing this CVE and the companion CVE-2026-51773 has appeared and should be treated as a possible proof-of-concept.
Sensitive file disclosure in Apache Roller 6.1.5 allows an authenticated weblog administrator (CVSS PR:L) to read arbitrary files on the application classpath - including Roller configuration files that may contain secrets - by authoring a Velocity template that uses the #include or #parse directive to load a resource outside the active theme namespace. Roller's Velocity sandbox does not confine these directives, no non-default configuration is required, and any weblog whose administrator can author templates is affected; no public exploit identified at time of analysis.
An insecure direct object reference in InvoicePlane's Users::change_password() endpoint (prior to 1.7.2) lets an authenticated secondary administrator replace the password of any other user account - most critically the primary administrator - without knowing or supplying the current password, resulting in full takeover of that account and its privileges. Exploitation requires an existing administrator-level session (CVSS PR:H) and a deployment with at least one other admin account to target, so this is a privilege-consolidation/lateral-movement flaw rather than an unauthenticated remote takeover; no public exploit code was identified at time of analysis and the issue is not confirmed as actively exploited. Vendor-released patch: 1.7.2.
Authenticated employees can read other employees' payroll records in Horilla HR 1.5.0-85 and earlier by substituting the emp_id, allowance_id, or deduction_id identifiers on the allowances_deductions_tab, view_single_allowance, and view_single_deduction endpoints in payroll/views/component_views.py. Exploitation requires only a valid logged-in account at any privilege level - no payroll-view permission and no ownership of the target record are needed - and exposes salary structures, allowance and deduction amounts, personal loan disbursements, and repayment schedules, making this a confidentiality-only disclosure flaw rated CVSS 6.5. There is no confirmed active exploitation (no CISA KEV listing) and no public exploit identified at time of analysis, but the absence of a complete fixed version leaves defenders with limited options.
Authenticated SQL injection in the PESI service of IBM Guardium Data Protection 12.2 allows a logged-in attacker with network reachability to the service to inject SQL through a parameter and read sensitive records from the product's internal database. The assessed vector (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N) indicates high confidentiality impact with no integrity or availability effects, and the prerequisite is a valid Guardium session - there is no pre-authentication path. No public exploit code has been identified at time of analysis and the issue is not confirmed as actively exploited; IBM has published a vendor advisory covering the fix, making this a genuine but conditional priority that is most urgent for internet- or broadly reachable PESI deployments with many low-privilege accounts.
Cross-site request forgery in InvoicePlane versions prior to 1.7.2 lets an attacker force a logged-in administrator's browser to submit forged POST requests that permanently delete financial and configuration records - including payments, recurring invoices, client-user assignments, users, email templates, families, and import data. The destructive delete()/stop() methods omit the ensure_valid_post_request() CSRF token check that other parts of the application rely on, so although the routes accept only POST, they never validate the request's token; exploitation requires the victim administrator to visit attacker-controlled web content while holding a valid session (CVSS 3.1 6.5, AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:N). No public exploit code or confirmed active exploitation was identified at time of analysis, and the integrity impact is bounded to records the admin session could already delete - there is no data exfiltration, and deleted records may be recoverable from backups.
Cross-site request forgery in InvoicePlane versions prior to 1.7.2 lets an attacker silently stop a victim's recurring invoices by luring an authenticated administrator into loading attacker-controlled content that auto-requests the state-changing GET endpoint /invoices/recurring/stop/{id}. Because the route performs a privileged action without CSRF token validation on a simple GET, the attacker needs only a known or brute-forceable recurring-invoice identifier and no user interaction beyond the admin having a live session. The impact is integrity-only and confined to interrupting recurring billing, which can cause financial loss for the organization; the vendor has confirmed the flaw (GHSA-qf9q-2hxm-4wh9) and fixed it in release 1.7.2. No public exploit code has been identified and the issue is not listed as actively exploited in CISA KEV, though the low attack complexity and zero-click nature of the request keep the practical risk moderate.
Authenticated attackers holding low-privileged credentials on an IBM Financial Transaction Manager (FTM) 4.x for Red Hat OpenShift deployment can exploit an XML external entity (XXE) injection flaw to disclose sensitive files and internal resources readable by the application, rated CVSS 3.1 base 6.5 (MEDIUM) with high confidentiality impact and no integrity or availability effect. Exploitation requires network reachability to an FTM endpoint that parses attacker-supplied XML with external-entity resolution enabled, plus valid credentials (PR:L) and no user interaction; because FTM is typically deployed as an internal back-office payments/financial processing system rather than an internet-facing service, exposure is usually limited to authenticated users and adjacent network positions. There is no public exploit identified at time of analysis and no evidence of confirmed active exploitation, and IBM has published an advisory for the affected versions.
Authenticated SQL injection in the Better Messages WordPress plugin (versions up to and including 3.0.4) allows subscribers and higher-privileged users to extract sensitive database information via the 'group_id' parameter, but only when the BuddyBoss Platform plugin is installed with its Social Groups component disabled and the wp_bp_groups table remains present. The vulnerability has a base CVSS score of 6.5 (Medium) with high confidentiality impact but no integrity or availability impact; the independent assessment rates attack complexity as High (AC:H) due to the specific configuration prerequisite. No public exploit code or active exploitation has been identified at time of analysis.
Incomplete property masking in Brocade SANnav's logging subsystem can cause SNMP authentication and privacy passwords to be written into application logs whenever specific configuration conditions are met, so anyone who can read those logs or a generated support bundle can recover the credentials and use them for unauthorized read or management access to the monitored switch environment. The flaw is not remotely exploitable by an unauthenticated attacker: per the available CVSS data (PR:L) and our independent assessment (CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N), exploitation requires an actor who already holds log or support-bundle read privileges, and the sensitive values only appear under non-universal configuration conditions. No public exploit code and no confirmed active exploitation were identified at time of analysis; the practical risk is a credential-hygiene and information-disclosure exposure realized by insiders or by a post-compromise actor who reuses the recovered SNMP credentials against reachable switches.