CVE Database

Dell ThinOS 10, versions prior to 2605_10.2616, contain a Download of Code Without Integrity Check vulnerability. An unauthenticated attacker with physical access could potentially exploit this vulnerability, leading to arbitrary code execution.

Dell ThinOS 10, versions prior to 2605_10.2616, contain a Security Version Number Mutable to Older Versions vulnerability. A low privileged attacker with physical access could potentially exploit this vulnerability, leading to Protection mechanism bypass.

Dell ThinOS 10, versions prior to 2605_10.2616, contain a Missing Support for Integrity Check vulnerability. A high privileged attacker with local access could potentially exploit this vulnerability, leading to arbitrary code execution.

AppFlowy-Cloud versions 0.7.2 through 0.9.64 fail to authorize callers against the workspace in the bulk publish endpoint path, allowing authenticated users to publish content into other tenants' namespaces. Attackers can write published views with attacker-controlled title, body and metadata into victim workspaces to deface public pages or host phishing content on trusted URLs.

Flextype CMS through 1.0.0-alpha.3 accepts API authentication credentials through URL query string parameters in REST API routes. Attackers with access to web server, proxy, or monitoring logs can recover valid API token pairs that grant full API access.

Open WebUI is an extensible, feature-rich, and user-friendly self-hosted AI platform. From 0.8.0 until 0.11.1, Open WebUI's OAuth token exchange endpoint issues a session for a provider access token without running the OAuth role management that the normal OAuth login callback runs. A user whose provider roles the login callback would refuse, or would demote, could still obtain a working session at their existing role through this endpoint. This issue is fixed in version 0.11.1.

Open WebUI is an extensible, feature-rich, and user-friendly self-hosted AI platform. From 0.8.0 until 0.9.0, Open WebUI's OAuth token exchange endpoint issues a session for a provider access token without applying the email domain allowlist that the normal OAuth login callback enforces. An account whose email domain the login callback would refuse could still obtain a working session through this endpoint. This issue is fixed in version 0.9.0.

import_contacts Path Traversal in Groundhogg <= 4.7.1 versions.

Unauthenticated Broken Access Control in Salon booking system <= 10.31.5 versions.

Subscriber Cross Site Scripting (XSS) in EventON <= 2.5.7 versions.

Subscriber Broken Access Control in IMPress for IDX Broker <= 3.3.0 versions.

Unauthenticated Broken Authentication in IMPress for IDX Broker <= 3.3.0 versions.

Unauthenticated Broken Access Control in BuddyForms <= 2.9.0 versions.

Subscriber Cross Site Scripting (XSS) in WP Docs <= 2.3.1 versions.

Subscriber Arbitrary File Download in Youzify <= 1.3.7 versions.

Unauthenticated Broken Access Control in Robokassa payment gateway for Woocommerce <= 1.8.9 versions.

Unauthenticated Bypass Vulnerability in Simple Cloudflare Turnstile <= 1.42.1 versions.

Unauthenticated Content Injection in Simple Cloudflare Turnstile <= 1.42.1 versions.

The Sierra Wireless HL78xx modem GNSS driver (drivers/modem/hl78xx/, later drivers/modem/vendor_standalone/hl78xx/) embeds a generic struct gnss_nmea0183_match_data match_data inside struct hl78xx_gnss_data. The generic NMEA0183 match helper (drivers/gnss/gnss_nmea0183_match.c) requires that context to be the first member because its callbacks cast user_data directly to struct gnss_nmea0183_match_data . In the affected releases match_data was the second member (after const struct device dev), so it sat at a non-zero offset while gnss_nmea0183_match_init() initialized it at the correct address. The registered NMEA handlers instead pass the whole device data object (data->devices.gnss->data, offset 0), producing an offset-shifted type confusion between where state is initialized and where the parse callbacks read and write it. When NMEA sentences from the GNSS receiver are parsed, the GGA/RMC callbacks write parsed fix data into the wrong location within the struct, and the GSV callback (gnss_nmea0183_match_gsv_callback, active under CONFIG_GNSS_SATELLITES) reads its satellites pointer and bound from the wrong offsets — non-pointer bytes of struct hl78xx_gnss_data — and then writes parsed struct gnss_satellite entries through that bogus pointer. This is a write through an uninitialized/wild pointer with a garbage bound. The NMEA handlers are registered by default (CONFIG_HL78XX_GNSS_SOURCE_NMEA is the default GNSS source) on devices using the HL78xx GNSS. The driver runs in kernel context and the NMEA data originates from the GNSS radio front-end, so a party able to influence the GNSS signal (for example GNSS/GPS spoofing at radio proximity) can drive the kernel-side parser into the faulty write. The most likely impact is a crash (denial of service) because the bogus pointer resolves to a fixed near-NULL value, with adjacent-memory corruption possible on MMU-less targets. Confidentiality is not affected. Exploitation requires the satellites feature to be enabled and active, so attack complexity is high.

MISP contains an HTML injection vulnerability in the MISPElementHTMLFormatterTool component, which is responsible for rendering MISP element references (attributes, objects, and tags) into inline HTML during PDF report export via the convert_markdown_to_pdf module. The attribute(), objectAttribute(), object(), and tag() methods interpolated user-controlled fields (attribute type, attribute value, object name, object relation, tag name, tag colour, and tag text colour) directly into HTML templates without applying HTML entity encoding. An authenticated user with the ability to create or modify MISP attributes, objects, or tags could embed arbitrary HTML markup in these fields. When a report containing such elements was exported to PDF, the unescaped content was rendered as live HTML rather than inert text, potentially injecting script tags, breaking the document structure, or altering the visual content of the exported report. Additionally, the attribute() method contained a template with hardcoded sample values ("domain-ip" and "google.com") instead of format placeholders, meaning every plain attribute reference in a PDF displayed the sample text rather than the actual indicator value, constituting a data-integrity defect in the exported document. The vulnerability requires an authenticated actor with write access to MISP elements and a subsequent PDF export of a report referencing those elements. The security impact is primarily to the integrity of the exported document and, depending on the HTML-to-PDF rendering engine, potential execution of injected markup during the conversion step. Version affected: ≤2.5.45

EspoCRM before 10.0.4 is vulnerable to server-side request forgery. HostCheck::ipAddressIsNotInternal(), which validates outbound URLs to block requests to internal/private IP addresses, strips ::ffff: (IPv4-mapped IPv6) prefixes but does not recognize IPv6 transition addresses that embed private IPv4 addresses: NAT64 (64:ff9b::), 6to4 (2002::), and Teredo (2001:0000::). An attacker who controls a domain with AAAA records pointing to such transition addresses can bypass both the internal-host validation and the CURLOPT_RESOLVE IP-pinning check, causing EspoCRM to issue outbound requests to internal network services. Affected paths include POST /Attachment/fromImageUrl, reachable by any authenticated user with attachment access, and outbound webhook delivery, which requires an admin or API user.

Snipe-IT's predefined kit checkout path does not enforce Full Multiple Company Support (FMCS) tenant isolation on the checkout target. Unlike the single, bulk, API, accessory, license and consumable checkout paths, App\Services\PredefinedKitCheckoutService never calls $item->canCheckoutTo($target); it only performs the actor-vs-item policy check and an availability check before persisting the checkout. With FMCS enabled, a non-superuser who belongs to at least two companies and holds the assets.checkout permission can POST to /kits/{kit}/checkout with a user_id belonging only to company B and have a company-A asset (and likewise kit licenses, consumables and accessories) assigned to that user, bypassing the company-mismatch check that blocks the same operation on every other checkout path. The issue is fixed in Snipe-IT 8.7.2; it was runtime-verified on v8.6.3 and code-inspected on v8.7.1, and the affected service has lacked the check since 2019, so earlier FMCS deployments are likely also affected.

OpenPanel is an analytics platform. In all versions (no patched release available at time of publication), the data importer fetches a caller-supplied URL with plain fetch instead of the project's existing SSRF guard (apps/api/src/utils/safe-fetch.ts). In packages/importer/src/providers/umami.ts, parseRemoteFile calls fetch() on config.fileUrl, which is validated only by z.string().url(), so values such as http://127.0.0.1:9911/ or http://169.254.169.254/latest/meta-data/ are accepted; the shared createFileImportConfig factory gives the plausible provider the same field. An authenticated organization member — including a default 'member' with no project_access rows, for whom the intended access-level check is skipped because getProjectAccess returns boolean true rather than a level object — can therefore make the server connect to any address reachable from it. The resulting HTTP status and status text are persisted as Import.errorMessage and returned by import.get to the same user, providing a scanning oracle for internal hosts, ports and paths; if an internal response parses as Umami CSV, its rows are ingested as events and become readable in the attacker's analytics views.

Renovate is a dependency update automation tool. In versions before 44.3.1 (and Mend Renovate CE/EE images before 15.4.0, mend-renovate-ce Helm chart before 15.4.0, mend-renovate-enterprise-edition Helm chart before 10.4.0), digest updates are not subject to the internal `minimumReleaseAge` (stability age) checks. When a repository configures `minimumReleaseAge` and has dependencies with `updateType=digest` — for example GitHub Actions pinned to a commit SHA with a floating tag, Docker images, Go modules or NuGet packages — Renovate will still open a pull request for a newly published digest, marked only with a pending `renovate/stability-days` status check. A newly published, potentially malicious dependency version can therefore cause a PR to be raised and CI workflows to potentially run before the configured minimum release age has elapsed, which is precisely what the Minimum Release Age control is intended to prevent. The issue is fixed in Renovate 44.3.1; as a workaround, digest updates can be disabled or gated behind `dependencyDashboardApproval`.

Traefik is an HTTP reverse proxy and load balancer. In versions >= v2.8.2 through <= v2.11.55 and >= v3.0.0 through <= v3.7.11, the entryPoints.<name>.transport.respondingTimeouts settings — notably readTimeout, which is enabled by default at 60s — are not applied to the HTTP/3 request path. readTimeout is enforced as a deadline on the underlying TCP connection, which cannot be applied to a QUIC stream, and Traefik's HTTP/3 server is constructed without any timeout. As a result, on entry points with HTTP/3 enabled, an unauthenticated remote client that trickles request body bytes can hold a request open indefinitely and, with it, one upstream connection per request, exhausting bounded backend connection pools and causing denial of service. The issue was introduced in v2.8.2 when a quic-go API change removed the embedded http.Server that carried these timeouts. Fixed in v2.11.56 and v3.7.12.

AVideo through revision c3edcc274c389816d434acadac07ee78eaf330c1 (master, 2026-08-23) incompletely sanitizes sensitive user fields in the APIName=video response. Video rows include columns joined from the video owner's user record, and API::get_api_video() calls removeSensitiveUserFields() only when the caller is neither authenticated nor using a valid API secret; it never compares User::getId() with the row's users_id. As a result, any low-privilege authenticated account can request another user's video and receive the owner's email address, account name (user), and isAdmin, canUpload and canStream flags, which are removed for guests and should be returned only to the owner or to callers with a valid API secret. No patched version was available at the time of the advisory.

WWBN AVideo through commit c3edcc274c389816d434acadac07ee78eaf330c1 (master, 2026-08-23) contains a cross-site request forgery vulnerability in the CustomizeUser plugin's plugin/CustomizeUser/setSubscribers.json.php endpoint. The script reads users_id and ExtraSubscribers from $_REQUEST and calls User::setExtraSubscribers() without requiring a POST request or a valid CSRF token; autoCSRFGuard() is applied only to POST requests to *.json.php, and the mutation does not pass through ObjectYPT::save()/isUntrustedRequest(), so a cross-origin GET succeeds. An unauthenticated remote attacker who induces a logged-in administrator to load a crafted page (for example an <img> tag pointing at the endpoint) can forge the ExtraSubscribers count displayed on any user's channel. The CustomizeUser plugin is enabled by default, and no patched version was available at the time of reporting.

Capgo fails to clean up channel permission overrides when a user's last organization role binding is deleted, leaving stale overrides active. Attackers can retain channel-specific permissions after their base RBAC access has been revoked to perform unauthorized actions like changing production OTA versions.

A security vulnerability has been detected in proma-ai Proma up to 0.19.37. Affected is the function resolveTargetPath of the file apps/electron/src/main/lib/file-preview-service.ts of the component File Preview Service. Such manipulation of the argument file_path leads to path traversal. Local access is required to approach this attack. The exploit has been disclosed publicly and may be used. Upgrading to version 0.19.52 is able to address this issue. The name of the patch is b7bf78ab74b1552c92fc98c7db9a8a8d92c631df. It is suggested to upgrade the affected component. The PoC's candidateBasePaths parameter name does not match the current IPC API (0.19.52 uses FileAccessOptions with internal getPreviewCandidateBasePaths()). The reporter likely targeted 0.16.3 where the IPC handler accepted raw string[] base paths. The core vulnerability - basename-collision fallback - is independent of the parameter name and is confirmed in source.

Suricata is a network Intrusion Detection System, Intrusion Prevention System and Network Security Monitoring engine. Starting in version 8.0.0 and prior to version 8.0.5, when Lua rule execution is enabled, the Lua sandbox memory limit was not consistently enforced for new allocations. Certain Lua allocation patterns could exceed `security.lua.max-bytes` without triggering the intended memory limit, making the configured sandbox limit unreliable. This requires Lua rules to be enabled and an affected Lua script/rule to be loaded. Version 8.0.5 contains a fix. As a workaround, disable `security.lua.allow-rules` unless Lua rules are required.