Missing authorization in JeecgBoot 3.9.5 and earlier allows any authenticated low-privileged user to remove arbitrary users from any tenant's product pack by sending attacker-controlled userId and packId values to PUT /sys/tenant/deleteTenantPackUser, silently revoking assigned permissions such as tenant administrator access. Exploitation requires only a valid low-privileged account (PR:L per the assessed CVSS 3.1 vector AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:L/A:L) on an instance running in multi-tenant mode with tenant product packs configured, and publicly available exploit code exists, though there is no confirmed active exploitation (CISA KEV). The impact is confined to integrity and availability of permission assignments - no confidentiality exposure and no code execution - making this a genuine but moderate-priority issue rather than a top-tier emergency.
JeecgBoot through 3.9.5 contains an authorization bypass (CWE-639) in the AIRAG video-generation module, where an authenticated low-privileged user can read other users' AI video-generation history by supplying an arbitrary userId to the getVideoRecords handler in VideoGenerationController. Publicly available exploit code exists, and the disclosed records include prompts, task IDs, video URLs, and cover URLs; however, there is no write access or broader compromise, and exploitation requires the AIRAG video feature to be deployed and the attacker to already hold a valid account. The issue is a moderate-severity information-disclosure bug (CVSS 4.0 5.3; independent CVSS 3.1 assessment 5.3) rather than a top-tier priority.
Improper authorization in Strapi 5.47.0 through 5.57.0 allows authenticated holders of pre-existing admin API tokens to retain all-field Content Manager read access after an administrator field-restricts the token owner's role, because the reconciliation routine skips token permissions whose field entries are omitted or null. Publicly available exploit code exists, but exploitation requires a valid admin API token and only exposes read access to content fields, with no write, integrity or availability impact. CISA KEV does not list this issue and the assessed severity is moderate (CVSS 3.1 5.3, AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N).
Authenticated partners in Immich through 3.3.1 can read metadata from assets stored in another user's Locked Folder - including GPS coordinates, capture times, descriptions and camera details - via the partner synchronization stream, because sync queries fail to exclude assets marked with Locked visibility. The attacker must already hold an active, victim-accepted partner relationship and be authenticated, and publicly available exploit code exists demonstrating the technique (no CISA KEV confirmation of active exploitation). Impact is confined to metadata disclosure of Locked Folder assets; image/video binaries, credentials, and data belonging to non-partnered users are not exposed, making this a moderate-severity information-disclosure issue rather than an emergency.
Missing authorization in JeecgBoot through 3.9.5 lets any authenticated low-privileged user read arbitrary users' role codes by sending GET requests to /sys/api/queryUserRoles with an arbitrary username parameter, allowing role-assignment discovery and identification of administrator accounts. The issue is tracked as CWE-862 (missing authorization) with an assessed CVSS of 5.3 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:N/C:L/I:N/A:N), requiring only a valid JeecgBoot account of any privilege level; impact is confined to information disclosure rather than write or execution capability. Publicly available exploit code exists for this endpoint, but there is no indication of confirmed active exploitation (CISA KEV); the endpoint is exploitable in default deployments with no special configuration, making the authentication requirement the primary limiting factor.
Missing authorization in the JeecgBoot SysAnnouncementController downLoadFiles handler lets any authenticated low-privileged user retrieve a ZIP of announcement attachments by supplying a valid announcement id, exposing files from unpublished announcements or announcements addressed only to other users. The flaw affects JeecgBoot through version 3.9.5 (CPE cpe:2.3:a:jeecgboot:jeecgboot:*), requires a valid non-admin account and knowledge or enumeration of an announcement id, and yields only confidentiality impact with no write, execution, or availability effect. Publicly available exploit code exists (a Python PoC script published on GitHub) but there is no confirmed active exploitation (not listed in CISA KEV), so this is a medium-severity information-disclosure issue rather than a critical priority.
Missing authorization in JeecgBoot through 3.9.5 allows any authenticated low-privileged user to read OpenAPI gateway definitions they are not entitled to see, exposing internal origin URLs, virtual paths, IP whitelists, and header and parameter templates. The flaw (CWE-862) lives in the OpenApiController queryById handler, which fails to enforce the openapi permission, so a simple GET /openapi/queryById with a valid entry id returns configuration metadata. It is a moderate-severity issue (CVSS 5.3, PR:L, confidentiality-only impact) for which publicly available exploit code exists, but there is no indication of confirmed active exploitation and the CVE is not in CISA KEV.
Missing authorization in JeecgBoot through 3.9.5 lets any authenticated user - including low-privileged accounts - download every user's stored AI prompts by calling the /airag/prompts/exportXls endpoint. The response is an Excel workbook exposing prompt content, model IDs, and parameters belonging to all users, a moderate-severity information disclosure rather than a critical code-execution issue. Publicly available exploit code exists (a PoC script targets this handler), though there is no confirmed evidence of active exploitation (no CISA KEV listing), and exploitation is gated solely on possession of any valid low-privilege session.
Missing authorization in JeecgBoot's AiRAG knowledge module lets low-privileged authenticated users read document text chunks from knowledge bases they do not own. The embeddingSearch handler in AiragKnowledgeController (GET /airag/knowledge/embedding/search) lacks Shiro permission annotations, so any valid account can pass arbitrary knowledge base IDs and retrieve indexed content belonging to other users or tenants. Exploitation is remote and requires authentication (CVSS:4.0 PR:L, assessed CVSS:3.1 AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N), impact is limited to confidentiality, and publicly available exploit code exists (a Python PoC targeting the data plane); no confirmed active exploitation has been reported. This is a genuine but moderate-severity issue, not a top-tier emergency.
JeecgBoot versions through 3.9.5 contain a missing authorization flaw in the AiragPromptsController that lets any authenticated user read AI prompt templates belonging to other users, including administrators. Attackers can enumerate prompt IDs through the unguarded /airag/prompts/list endpoint and then retrieve each template's text, model parameters, and creator details via queryById, resulting in unauthorized information disclosure. Publicly available exploit code exists, but the issue requires a valid account and the AIRAG module to be deployed, making it a modest-severity confidentiality risk rather than a critical one.
An authenticated low-privileged user on JeecgBoot through 3.9.5 can query any tenant's product-pack configuration by calling the SysTenantController listPackByTenantUserId handler with attacker-chosen tenantId and userId parameters, because that handler performs no authorization check on the caller's ownership of the requested tenant. The result is read-only disclosure - tenant pack settings plus the identities of users who are tenant administrators - which is useful for reconnaissance and privilege-mapping but causes no data modification or service disruption (CVSS 5.3 / CVSS:3.1 AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N). Publicly available exploit code exists in a GitHub proof-of-concept repository, and exploitation is meaningful only in multi-tenant deployments where other tenants exist to enumerate; there is no confirmation of active exploitation (CISA KEV) at time of analysis.
Missing authorization in JeecgBoot's SysTenantController queryTenantAuthInfo handler (all versions through 3.9.5) allows any authenticated, low-privileged user to enumerate small integer tenant IDs and read other tenants' full sys_tenant records, including house numbers used as tenant join codes and company profile fields. Exploitation requires a valid authenticated account (CVSS vector PR:L) on a multi-tenant deployment where other tenants' records exist; publicly available exploit code exists, but there is no confirmation of active exploitation. The impact is limited to read-only information disclosure (CVSS 5.3, assessed as C:L/I:N/A:N), making this a moderate access-control issue rather than a high-urgency flaw.
Information disclosure in JeecgBoot through 3.9.5 allows any authenticated low-privileged user to read tenant administrator applications for any tenant by iterating the tenantId parameter on the /sys/tenant/getTenantPackApplyUsers endpoint. Public exploit code exists, but exploitation requires a valid account and a multi-tenant deployment with pending applicants; the impact is limited to read-only disclosure of applicant PII (usernames, real names, phone numbers, departments) with no write or availability effect. This medium-severity issue is confirmed by an independent assessment as an information-disclosure flaw, not a critical priority.
Missing authorization in JeecgBoot through 3.9.5 lets any authenticated, low-privileged user send spoofed template announcements by POSTing to /sys/api/sendTemplateAnnouncement with attacker-chosen sender, recipient, title, and template parameters, so the resulting messages appear to come from admin or system accounts. Rated 5.3 (medium), the issue carries only low integrity impact with no confidentiality or availability effect and no path to code execution, and it requires the attacker to already hold a valid JeecgBoot session. Publicly available exploit code exists (VulnCheck-reported POC plus a GitHub PoC script), though there is no indication of confirmed active exploitation.
JeecgBoot through 3.9.5 allows any authenticated low-privileged user to send spoofed system announcements via the /sys/api/sendBusAnnouncement endpoint because SystemApiController does not enforce the required permission check. An attacker can forge the sender, recipients, title, and content to impersonate administrators or the system, enabling phishing against other users. No code execution, data disclosure, or service disruption is possible, and publicly available exploit code exists; this is a low-severity authorization flaw rather than a remotely exploitable unauthenticated compromise.
Missing authorization in the SysAnnouncementController editIzTop handler of JeecgBoot through 3.9.5 allows any authenticated user at a low privilege level to pin or unpin system-wide announcements by sending a POST or PUT request carrying an arbitrary announcement id. The root cause is a pure access-control failure (CWE-862), not a memory or injection flaw, so the practical impact is limited to tampering with the pinned-notice banner displayed to all users, with no data disclosure, code execution, or service disruption. Publicly available exploit code exists and the issue was reported by VulnCheck, but the assessed severity is modest (CVSS 3.1 score 5.3, PR:L) because an attacker must already hold valid credentials and the resulting integrity change is low.
Missing authorization in JeecgBoot's AIRAG prompts module (through version 3.9.5) lets any authenticated, low-privileged user invoke the promptExperiment handler in AiragPromptsController and launch AI prompt evaluation runs against prompt templates and datasets that belong to other users. Because the handler performs no ownership or permission check on the supplied prompt-template and dataset IDs, an attacker who can enumerate or guess those identifiers can write result rows into the airag_ext_data table and alter dataset status, an integrity-only impact with no confidentiality or availability loss. Exploitation requires a valid low-privileged account plus an instance where the AIRAG module is deployed with a working LLM backend; publicly available exploit code exists, but there is no confirmed active exploitation (no CISA KEV listing) and the issue is rated moderate rather than critical.
Missing authorization in JeecgBoot through 3.9.5 allows a low-privileged authenticated user to undo administrator deletions of AI prompt templates by sending crafted PUT requests to the /airag/prompts/revertRecycleBin endpoint with chosen template IDs, clearing the deleted flag and returning them to active use. Because the endpoint performs no permission check, any account with a valid session - not just an administrator - can restore recycled content; publicly available exploit code exists, though exploitation still requires credentials (CVSS PR:L) and knowledge of valid deleted-template IDs, and the impact is limited to an integrity change on prompt template records with no confidentiality or availability effect. Per our independent assessment this is a genuinely low-severity, low-priority issue: it is not exploitable by unauthenticated attackers, requires no user interaction, and cannot be used to read data, escalate privileges, or disrupt the platform.
Tampering with AI prompt templates in JeecgBoot through 3.9.5 is possible for any authenticated user due to missing authorization in the AiragPromptsController edit handler. A low-privileged attacker can send PUT or POST requests to /airag/prompts/edit with a target template id to overwrite prompt text and model parameters created by administrators or other users, resulting in integrity-only impact. Publicly available exploit code exists, though no active exploitation has been confirmed; CVSS 5.3 (medium).
Missing authorization in JeecgBoot through 3.9.5 lets any authenticated, low-privileged user delete pending tenant-administrator applications belonging to other tenants via the SysTenantController deleteApply handler, which also sends rejection notifications to the affected applicants. Publicly available exploit code exists, though exploitation requires a valid low-privileged account plus valid or enumerable tenantId, packId and userId values for a pending application. Impact is limited to integrity of the tenant-onboarding workflow with no data disclosure or service outage, making this a genuine but moderate issue rather than a top-priority one.
Missing authorization in JeecgBoot (all versions through 3.9.5) lets any authenticated, low-privileged user alter tenant administrator application status for arbitrary tenants by sending a PUT request to /sys/tenant/updateApplyStatus with a tenant id they do not own. The endpoint performs no ownership or role check on the supplied identifier, so an ordinary account - including one obtained through self-registration - can toggle the applyStatus field and thereby enable or disable tenant-admin applications across the entire instance. Publicly available exploit code exists (a Python PoC is published in a public GitHub repository), though this is not confirmed as actively exploited in the wild; EPSS and KEV data were not provided, and the independent assessment rates this as a moderate-severity broken access control issue with limited blast radius, since impact is confined to status toggling rather than full tenant takeover.
Authentication bypass in iFlytek AstronRPA through 1.1.6 allows unauthenticated, network-positioned attackers to reach the /api/resource/ and /api/rpa-ai-service/ API route families by presenting an arbitrary Bearer token, because the OpenResty gateway's auth_handler.lua accepts any value without validating it. Beyond simply getting past the gate, an attacker can spoof the X-User-Id or user_id header to impersonate any user of the platform, yielding low-level read and write access to those API routes. There is no public exploit identified at time of analysis, the issue is not confirmed actively exploited (no CISA KEV listing), and no vendor-released patch has been identified at time of analysis; the practical impact ceiling is the two affected route families on a gateway reachable by the attacker, not full system compromise.
AniWorld Downloader versions before 5.3.0 ship a WebUI whose /login POST handler permits unlimited, unthrottled password guessing and leaks which usernames exist, letting unauthenticated remote attackers enumerate accounts and brute-force credentials on any instance whose web interface is reachable. The exposure is a compound weakness rather than a single primitive: verify_user in the web database layer only performed a password hash comparison when a matching user row existed, so the differing response time distinguished valid from invalid usernames, while the absence of any per-IP or per-account failure cap allowed unlimited attempts against the confirmed names. This is a genuine but moderate-severity weakness, not a critical priority - exploitation still requires the WebUI to be network-exposed (a local-only or firewalled deployment is not reachable) and successful takeover still depends on a guessable or weak password, so strong unique credentials blunt the attack. No confirmed active exploitation (CISA KEV) and no public exploit identified at time of analysis; the underlying issue and its fix are documented in the upstream issue tracker and a public technical write-up, and the vendor patched the flaw in v5.3.0.
Missing authorization in VillaTheme GIFT4U (gift4u-gift-cards-all-in-one-for-woo) versions 1.1.3 and earlier allows unauthenticated remote attackers to tamper with limited gift-card-related data on WordPress/WooCommerce sites where the plugin is active. The flaw (CWE-862) stems from the plugin failing to enforce an access-control check on a restricted function, and per the CVSS vector (PR:N/UI:N) no authentication or user interaction is required. No public exploit code or active exploitation has been identified at time of analysis, and the impact is limited to integrity (VI:L) with no confidentiality or availability impact, resulting in a medium-severity rating.
Server-side request forgery in PDFMathTranslate (pdf2zh) through 1.9.11 lets unauthenticated remote attackers force the server to fetch arbitrary URLs supplied through the Gradio web GUI's Link (translate-from-URL) input, reaching internal services and cloud metadata endpoints with the fetched content returned inside the generated PDF. The flaw exists because the translate_file handler forwards user-controlled URLs to download_with_limit without validating the URL scheme or destination address, and the CVSS vector confirms no privileges (PR:N) and no user interaction (UI:N) are needed. Exploitation is gated primarily by exposure: the Gradio interface must be network-reachable and able to route to internal or link-local addresses, and impact is confined to information disclosure (VC:L, no integrity or availability effect) - there is no confirmed active exploitation (not in CISA KEV) and no public exploit code identified at time of analysis.
Pingvin Share X before 1.22.0 ships with an ineffective authentication rate limiter, because throttler TTL values written in seconds are interpreted by NestJS as milliseconds - collapsing intended 60-second and 5-minute windows into roughly 60 ms and 5 ms. This allows unauthenticated remote attackers to send effectively unthrottled requests to /api/auth/signIn, /api/auth/signIn/totp and /api/auth/resetPassword, enabling brute-force attempts against user passwords and 6-digit TOTP codes; no public exploit code has been identified at time of analysis and there is no KEV entry, and the flaw is assessed as a moderate authentication-hardening failure rather than a direct code-execution or data-leak issue. Because the weakness only removes rate limiting, successful compromise still depends on guessing a valid password or a valid TOTP value within its short validity window, so strong unique passwords and the limited TOTP lifetime materially reduce real-world risk.
A rate-limiting bypass in Pingvin Share X before 1.22.0 lets unauthenticated remote attackers defeat per-IP throttling on the sign-in, TOTP, and password-reset endpoints by rotating spoofed X-Forwarded-For headers. Because backend/src/main.ts calls app.set('trust proxy', true) unconditionally, any client-supplied forwarding header is believed, so attackers can brute-force passwords or TOTP codes at unrestricted speed and poison IP-based logging with forged client addresses. Exploitation requires only network reachability of the auth API and the ability to send arbitrary headers - no special configuration - though real account compromise still depends on weak credentials or low-entropy codes. No public exploit code or confirmed active exploitation was identified at time of analysis, and vendor-patched version 1.22.0 is available (CWE-348, CVSS 4.0 base 6.9 / assessed CVSS 3.1 6.5).
Unbounded memory allocation in the ericmj/decimal Elixir library lets an attacker who controls the 'places' argument of Decimal.round/2 or Decimal.round/3 exhaust host memory and get the BEAM VM killed; all releases from 0.1.0 up to (but not including) 3.1.2 are affected. The function materialises the entire rounded result before the 34-digit default context precision is applied, so a single call such as Decimal.round(Decimal.new("1.5"), -50_000_000) allocates roughly 5.5 GB, while the oldest releases instead spin one loop iteration per decimal place and burn CPU in proportion to the argument. The official CVSS 4.0 base score is 6.9 (availability-only, VC:N/VI:N/VA:H), our independent assessment rates it CVSS 3.1 7.5 AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H, and no public exploit code or confirmed active exploitation was identified at time of analysis. Exposure is conditional: the vulnerability is only reachable when a consuming application forwards an unbounded, user-supplied scale or decimal-place count into round/2 or round/3 - the input limits previously added for CVE-2026-32686 explicitly do not cover the places argument.
Arbitrary file deletion in the WordPress plugin Presto Player Pro (versions 3.0.1 and earlier) lets an authenticated account with at least Author-level privileges supply a traversal path to a plugin file-handling function and delete files outside the intended scope, including files elsewhere in the WordPress installation. Patchstack scores it CVSS 6.8 with high availability impact and no confidentiality or integrity impact, and no publicly available exploit code or active exploitation was identified at time of analysis. The realistic exposure is gated by the privilege requirement: sites that do not host untrusted Author accounts or allow open author-tier self-registration are substantially less exposed.
Privilege escalation in the Bookly WordPress plugin (versions up to and including 28.4) allows authenticated attackers with subscriber-level access to overwrite the WordPress user binding of any Bookly staff record, including those owned by administrators, by manipulating the 'id' and 'wp_user_id' parameters in the staff-edit AJAX endpoint. Exploitation requires the attacker to have a subscriber account and to be linked to at least one Bookly staff record, a condition present by default on fresh installs where the bookly_gen_allow_staff_edit_profile option is enabled. No public exploit code has been identified and CISA KEV does not list this vulnerability; the assessed CVSS is 6.8 with high attack complexity (AC:H) due to the prerequisite of a staff-linked subscriber account.
Stored cross-site scripting in the Gallery PhotoBlocks WordPress plugin before 1.3.6 allows authenticated users with Contributor-level access or higher to save JavaScript into one of the plugin's gallery settings, which is then written into an HTML attribute without sanitization or escaping, so the payload executes in the browser of any user - including an administrator - who views a page containing the gallery. Exploitation is not self-triggering: an attacker needs at least a Contributor account plus a victim who loads the affected page (UI:R), and sites that do not grant untrusted users posting roles are largely unaffected. No public exploit code has been identified at time of analysis and no EPSS score was supplied in the available intelligence; the vendor has released a fix, and our independent reassessment rates the issue MEDIUM (CVSS 3.1 base reflecting PR:L, S:C, C:L/I:L/A:N), which is lower in privilege prerequisite but broader in scope than NVD's published 6.8 vector.
Local unprivileged users on a host running GenOffice's HTTP MCP server can read other users' uploaded documents and AI-converted outputs, because every version through 0.11.505 stages those files in a genoffice-mcp-http directory under the system temporary directory with world-readable permissions. The HTTP bearer token is irrelevant to the attack: an attacker simply lists and opens the temp-directory contents at the filesystem layer, so the bearer token that normally guards the MCP HTTP endpoint provides no protection at all. The issue is strictly local (CVSS 4.0 score 6.8, AV:L/PR:L with high confidentiality impact and no integrity or availability impact), read-only, and bounded to files currently staged in the temp directory; no public exploit code or active exploitation has been identified at time of analysis, and single-user hosts with no other local accounts are effectively not exploitable.
Authenticated directory traversal in the SiteOrigin Widgets Bundle WordPress plugin (all versions up to and including 1.74.3) allows contributor-level users and above to read arbitrary files on the server through the get_instance_css function, exposing sensitive data such as wp-config.php secrets. Exploitation requires only a low-privilege contributor account (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N), needs no user interaction, and is limited to read-only information disclosure with no direct write or code-execution impact; there is no public exploit identified at time of analysis and the issue is not confirmed actively exploited. Overall this is a credible but moderate-priority issue rather than a top-tier emergency, with risk concentrated on sites that grant contributor access broadly.
Missing authorization checks in the Jetpack VideoPress WordPress plugin versions 3.6 and earlier let an authenticated Subscriber-level account perform actions that should be limited to higher-privileged users, giving a low-privileged attacker the ability to modify data (integrity impact only, with no confidentiality or availability effect per the assessed CVSS vector CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:N, base score 6.5). Sites are exposed if they run an affected plugin version and allow the attacker to obtain a Subscriber account, which is trivial on installations with open user registration enabled; fully unauthenticated attackers cannot exploit the flaw. No public exploit code has been identified at time of analysis and the issue is not confirmed actively exploited (no CISA KEV listing), so this is a moderate-priority hardening item rather than an emergency, though the absence of a named vulnerable endpoint in the published advisory makes the true reachable surface hard to bound.
Authenticated SQL injection in WP Booking System - Booking Calendar versions up to and including 2.1.0.1 allows subscriber-level attackers to extract sensitive database information via the 'current_month' parameter. The vulnerability is a regression: a capability check added in 2.0.19.11-2.0.19.14 to fix CVE-2024-50425 was removed in version 2.1, and no nonce validation is required, so any logged-in subscriber can reach the vulnerable AJAX handler. No public exploit code or active exploitation has been identified at time of analysis, and the assessed severity is moderate (CVSS 6.5).
Time-based blind SQL injection in the WP Directory Kit plugin for WordPress (all versions up to and including 1.5.9) allows authenticated attackers holding subscriber-level access or above to append extra SQL into an existing query and read sensitive information from the site database. The issue is second-order: a subscriber stores a display_name value containing a single quote, which WordPress preserves verbatim, and the payload only fires later when WdkCachedUserEditor::update_listings_user_editor() re-reads that cached value into the SQL sink without escaping or prepared-statement parameterization. Rated CVSS 6.5 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N), the impact is confidentiality-only and blind, so exfiltration is slow and generates measurable request volume; no public exploit code has been identified at time of analysis and there is no confirmed active exploitation (CISA KEV).
Authorization bypass in the Academy LMS - AI Course Builder, Quizzes, Certificates & eLearning WordPress plugin (all versions up to and including 4.0.3) allows any user who holds an instructor role for at least one course to permanently delete lesson comments and their replies belonging to courses they do not instruct, including comments authored by administrators. Exploitation requires an authenticated account with instructor-level access (CVSS PR:L), and the attacker supplies their own valid course_id to pass the plugin's flawed instructor check while targeting comment IDs from other courses; the abuse results in irreversible content destruction (integrity impact High, confidentiality and availability unaffected per the assessed vector). No public exploit code has been identified at time of analysis, the issue is not listed as confirmed actively exploited (CISA KEV), and the assessed severity is moderate (CVSS 6.5) rather than an emergency, primarily because the instructor-account prerequisite confines the threat to semi-trusted internal users.
Unauthenticated content injection in the Kutethemes 'Boutique' WordPress theme at version 2.3.3 and earlier lets remote attackers reach a shortcode-handling path and inject arbitrary content into affected sites. No authentication or user interaction is required (CVSS:3.1/AV:N/AC:L/PR:N/UI:N), so any internet-facing WordPress installation with this commercial theme active is exposed by default; the attack surface is bounded to sites running this specific theme. Impact is assessed as limited content injection and partial data exposure (C:L/I:L/A:N) rather than server compromise, and no public exploit code has been identified at time of analysis, so the practical risk is concentrated in unpatched, theme-deployed sites.
Trusted-credential abuse in CodePeople2 Sell Downloads (WordPress plugin) up to version 1.2.3 allows unauthenticated remote attackers to bypass origin verification and access or modify limited data, such as download or order metadata. The flaw, tracked as CVE-2026-94590, has a CVSS score of 6.5 and is not known to be actively exploited; no public exploit code has been identified. Only WordPress sites with this specific, low-prevalence plugin active are affected, making the real-world risk moderate and low-priority.
Broken access control in the Tutor LMS WordPress plugin before 4.1.1 lets an authenticated user with instructor-level privileges reassign the parent of arbitrary posts on the site, allowing them to absorb other instructors' course content into their own courses and to orphan legitimate published content so it becomes unreachable. The root issue is an insecure direct object reference (CWE-639) in the plugin's course content ordering handler, which never verifies that the posts named in the request belong to a course the requester manages. No public exploit code has been identified at time of analysis, and the practical risk is confined to multi-instructor deployments that hand instructor roles to semi-trusted or self-registered users; the vendor has released a fix in Tutor LMS 4.1.1.
Stored cross-site scripting in Premium Packages - Sell Digital Products Securely (wpdm-premium-packages) through version 7.2.6 lets authenticated attackers with subscriber-level access or above inject arbitrary web scripts into checkout billing fields such as phone, state, taxid, and email, which are then rendered in the customer dashboard without adequate escaping. The issue is rated CVSS 6.4 Medium and requires an account on the target WordPress site, but exploitation does not require administrative privileges; script execution occurs when a higher-privileged user views the poisoned billing information. No public exploit has been identified at time of analysis, and the primary fix is an upstream changeset rather than a confirmed tagged patched release.
Stored DOM-based cross-site scripting in the SureDash - Community, Courses & Member Dashboard plugin for WordPress (all versions up to and including 1.12.1) allows authenticated attackers with subscriber-level access or above to plant arbitrary script inside community post content by entity-encoding a payload in an image's 'alt' attribute. The payload evades server-side wp_kses filtering because kses allows the img/alt combination and does not normalize HTML entities within attribute values, and it only executes client-side when the bundled GLightbox library reads the .alt DOM property and writes the decoded value into innerHTML as a victim views the injected page. There is no public exploit identified at time of analysis, and exploitation demands a valid low-privilege account plus an active community/GLightbox rendering path, so this is a genuine but moderate-priority issue rather than a top-tier emergency; the vendor has released SureDash 1.12.2 to address it.
Stored cross-site scripting in the WordPress plugin Data Tables Generator by Supsystic (all versions up to and including 1.15.1) allows an authenticated attacker holding a Subscriber-level or higher account to inject arbitrary JavaScript into a table cell's 'data' value through the updateRows action, which then executes in the browser of any user who views a page rendering that table. Reachability depends on a non-default precondition: an administrator must have deliberately added the attacker's low-privilege role to the plugin's 'access_roles' setting, which is a documented feature that hands those users the dtgs_nonce needed to call the vulnerable handler, so the exploitable population is meaningfully narrower than the raw base vector implies. No public exploit code has been identified at time of analysis and there is no KEV listing.
Stored cross-site scripting in the FV Flowplayer Video Player plugin for WordPress (FolioVision, all versions up to and including 7.5.54.7212) allows an authenticated attacker holding subscriber-level privileges or higher to inject arbitrary JavaScript through an unquoted popup shortcode attribute, which then executes in the browser of any user who later views the affected author bio/archive page. Exploitation is constrained by two prerequisites: the site must have the non-default profile_videos_enable_bio option enabled, and the attacker must first obtain a low-privileged account, which open registration makes feasible but which sites that disable public registration largely close off. At CVSS 6.4 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:N) this is a moderate-severity issue; no public exploit code was identified at time of analysis, so it should be handled as a patch-in-normal-cycle item rather than an emergency.
Stored cross-site scripting in the WP User Frontend WordPress plugin (all versions up to and including 4.3.12) allows any authenticated user with subscriber-level access or higher to persist arbitrary JavaScript that executes when a privileged user - typically an administrator - opens the affected transactions page. The root cause is insufficient sanitization and output escaping of the 'wpuf_payment_method' parameter: attackers bypass sanitize_text_field() by appending a valid gateway keyword such as 'bank' to the HTML payload, so the filter returns only the keyword for gateway routing while the full raw malicious value is stored and later rendered unescaped in the admin transactions list. Wordfence reported the issue with a CVSS 3.1 base score of 6.4 (AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:N); no public exploit code was identified at time of analysis, and there is no confirmed evidence of active exploitation (no CISA KEV listing).
Stored cross-site scripting in the Kadence Blocks - Page Builder Toolkit for Gutenberg Editor WordPress plugin affects all versions up to and including 3.7.12, allowing an authenticated attacker holding author-level privileges or higher to inject arbitrary web scripts through a block's font-family attribute that execute when a victim views the injected page. The payload path is conditional: the malicious font-family encoding bypass only activates when the non-default 'Load Google Fonts Locally' site option (kadence_blocks_font_settings['load_fonts_local']) is enabled, so default-configured sites are not exploitable through this vector. Exploitation additionally requires a victim to later access the poisoned page, and no public exploit code or confirmed active exploitation was identified at time of analysis; this is a moderate, conditionally-exploitable issue rather than a top-tier priority given the privileged account requirement, the non-default setting, and the necessary victim interaction.
Stored cross-site scripting in the Feedzy RSS Aggregator WordPress plugin (all versions up to and including 5.2.10) lets an authenticated attacker with contributor-level access or above inject arbitrary JavaScript into pages by pointing the Gutenberg Feedzy Loop Block at an attacker-controlled RSS feed whose <title> element contains malicious HTML, which is then rendered into the page without escaping. Because post_content stores only a benign reference to the external feed URL, the payload evades save-time wp_kses filtering entirely and executes for any user who views the affected page. Exploitation requires a contributor account, the Loop Block in use, and attacker control of an external feed; no public exploit code has been identified at time of analysis and the plugin is not listed in CISA KEV, so risk is bounded by these preconditions rather than by the network-facing vector (CVSS 6.4).
Stored cross-site scripting in the GiveWP donation and fundraising plugin for WordPress (versions up to and including 4.17.0) lets authenticated attackers with Contributor-level access or above plant arbitrary JavaScript through the givewp_campaign_grid shortcode's attributes, which is then rendered unsanitized into any page or preview that loads the affected content. Exploitation requires a low-privileged account able to submit shortcode-bearing content and a higher-privileged or public visitor to view it, which is why it rates only moderate (CVSS 6.4, scope-changed C:L/I:L) and is a realistically lower-priority issue; no public exploit code has been identified at time of analysis and there is no confirmed active exploitation (no CISA KEV listing). The primary impact is session/token theft or privileged action abuse against the viewing user rather than site-wide compromise.
Stored cross-site scripting in the Simple Membership WordPress plugin (versions up to and including 4.8.4) lets an attacker persist arbitrary JavaScript through the member profile 'country' field, which then executes in the browser of any user - typically an administrator - who views the affected member listing or profile page. By default the attacker must hold an authenticated account at subscriber level or above, but the vector widens to fully unauthenticated attackers whenever the plugin's 'Enable Free Membership' option is on, because that setting permits anonymous front-end registration and profile submission. Rated CVSS 6.4 (Medium) and reported by Wordfence, there is no public exploit identified at time of analysis and no confirmed active exploitation (not in CISA KEV); the fix is the vendor-released version 4.8.5.
Stored cross-site scripting (XSS) in the Retainful Email Marketing for WordPress and WooCommerce plugin (versions ≤1.0.10) allows authenticated attackers with subscriber-level access to inject malicious scripts via the 'data' parameter in the plugin's REST API, which execute when a victim views the affected page. This vulnerability requires a low-privilege account and user interaction, and no public exploit code or active exploitation has been identified. It is a moderate-severity (CVSS 6.4) issue that should be patched promptly.