| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| FastGPT is an AI Agent building platform. In 4.14.7 and earlier, FastGPT's Python Sandbox (fastgpt-sandbox) includes guardrails intended to prevent file writes (static detection + seccomp). These guardrails are bypassable by remapping stdout (fd 1) to an arbitrary writable file descriptor using fcntl. After remapping, writing via sys.stdout.write() still satisfies the seccomp rule write(fd==1), enabling arbitrary file creation/overwrite inside the sandbox container despite the intended no file writes restriction. |
| The grafanacubism-panel plugin allows use of cubism.js in Grafana. In 0.1.2 and earlier, the panel's zoom-link handler passes a dashboard-editor-supplied URL directly to window.location.assign() / window.open() with no scheme validation. An attacker with dashboard Editor privileges can set the link to a javascript: URI; when any Viewer drag-zooms on the panel, the payload executes in the Grafana origin. |
| ha-mcp is a Home Assistant MCP Server. Prior to 7.0.0, the ha-mcp OAuth consent form (beta feature) accepts a user-supplied ha_url and makes a server-side HTTP request to {ha_url}/api/config with no URL validation. An unauthenticated attacker can submit arbitrary URLs to perform internal network reconnaissance via an error oracle. Two additional code paths in OAuth tool calls (REST and WebSocket) are affected by the same primitive. The primary deployment method (private URL with pre-configured HOMEASSISTANT_TOKEN) is not affected. This vulnerability is fixed in 7.0.0. |
| StudioCMS is a server-side-rendered, Astro native, headless content management system. Prior to 0.4.3, the updateUserNotifications endpoint accepts a user ID from the request payload and uses it to update that user's notification preferences. It checks that the caller is logged in but never verifies that the caller owns the target account (id !== userData.user.id). Any authenticated visitor can modify notification preferences for any user, including disabling admin notifications to suppress detection of malicious activity. This vulnerability is fixed in 0.4.3. |
| StudioCMS is a server-side-rendered, Astro native, headless content management system. Prior to 0.4.3, the POST /studiocms_api/dashboard/create-reset-link endpoint allows any authenticated user with admin privileges to generate a password reset token for any other user, including the owner account. The handler verifies that the caller is an admin but does not enforce role hierarchy, nor does it validate that the target userId matches the caller's identity. Combined with the POST /studiocms_api/dashboard/reset-password endpoint, this allows a complete account takeover of the highest-privileged account in the system. This vulnerability is fixed in 0.4.3. |
| OliveTin gives access to predefined shell commands from a web interface. In 3000.10.2 and earlier, OliveTin’s live EventStream broadcasts execution events and action output to authenticated dashboard subscribers without enforcing per-action authorization. A low-privileged authenticated user can receive output from actions they are not allowed to view, resulting in broken access control and sensitive information disclosure. |
| OpenProject is an open-source, web-based project management software. Prior to 17.2.0, OpenProject SMTP test endpoint (POST /admin/settings/mail_notifications) accepts arbitrary host and port values and exhibits measurable differences in response behaviour depending on whether the target IP exists and whether the port is open. An attacker with access can use these timing and error distinctions to map internal hosts and identify which services/ports are reachable. Similarly, you can create webhooks in OpenProject and point them to arbitrary IPs, resulting in the same kind of SSRF issue which allows attackers to scan the internal network. This vulnerability is fixed in 17.2.0. |
| Emlog is an open source website building system. In 2.6.6 and earlier, the delete_async action (asynchronous delete) lacks a call to LoginAuth::checkToken(), enabling CSRF attacks. |
| Argo Workflows is an open source container-native workflow engine for orchestrating parallel jobs on Kubernetes. From 2.9.0 to before 4.0.2 and 3.7.11, A user who can submit Workflows can completely bypass all security settings defined in a WorkflowTemplate by including a podSpecPatch field in their Workflow submission. This works even when the controller is configured with templateReferencing: Strict, which is specifically documented as a mechanism to restrict users to admin-approved templates. The podSpecPatch field on a submitted Workflow takes precedence over the referenced WorkflowTemplate during spec merging and is applied directly to the pod spec at creation time with no security validation. This vulnerability is fixed in 4.0.2 and 3.7.11. |
| Taskosaur is an open source project management platform with conversational AI for task execution in-app. In 1.0.0, the application does not properly validate or restrict the role parameter during the user registration process. An attacker can manually modify the request payload and assign themselves elevated privileges. Because the backend does not enforce role assignment restrictions or ignore client-supplied role parameters, the server accepts the manipulated value and creates the account with SUPER_ADMIN privileges. This allows any unauthenticated attacker to register a fully privileged administrative account. |
| cpp-httplib is a C++11 single-file header-only cross platform HTTP/HTTPS library. Prior to 0.37.1, when a cpp-httplib client uses the streaming API (httplib::stream::Get, httplib::stream::Post, etc.), the library calls std::stoull() directly on the Content-Length header value received from the server with no input validation and no exception handling. std::stoull throws std::invalid_argument for non-numeric strings and std::out_of_range for values exceeding ULLONG_MAX. Since nothing catches these exceptions, the C++ runtime calls std::terminate(), which kills the process with SIGABRT. Any server the client connects to — including servers reached via HTTP redirects, third-party APIs, or man-in-the-middle positions can crash the client application with a single HTTP response. No authentication is required. No interaction from the end user is required. The crash is deterministic and immediate. This vulnerability is fixed in 0.37.1. |
| Craft Commerce is an ecommerce platform for Craft CMS. Prior to 4.11.0 and 5.6.0, An Insecure Direct Object Reference (IDOR) vulnerability exists in Craft Commerce’s cart functionality that allows users to hijack any shopping cart by knowing or guessing its 32-character number. The CartController accepts a user-supplied number parameter to load and modify shopping carts. No ownership validation is performed - the code only checks if the order exists and is incomplete, not whether the requester has authorization to access it. This vulnerability enables the takeover of shopping sessions and potential exposure of PII. This vulnerability is fixed in 4.11.0 and 5.6.0. |
| flagd is a feature flag daemon with a Unix philosophy. Prior to 0.14.2, flagd exposes OFREP (/ofrep/v1/evaluate/...) and gRPC (evaluation.v1, evaluation.v2) endpoints for feature flag evaluation. These endpoints are designed to be publicly accessible by client applications. The evaluation context included in request payloads is read into memory without any size restriction. An attacker can send a single HTTP request with an arbitrarily large body, causing flagd to allocate a corresponding amount of memory. This leads to immediate memory exhaustion and process termination (e.g., OOMKill in Kubernetes environments). flagd does not natively enforce authentication on its evaluation endpoints. While operators may deploy flagd behind an authenticating reverse proxy or similar infrastructure, the endpoints themselves impose no access control by default. This vulnerability is fixed in 0.14.2. |
| Cloud CLI (aka Claude Code UI) is a desktop and mobile UI for Claude Code, Cursor CLI, Codex, and Gemini-CLI. Prior to 1.24.0, The /api/user/git-config endpoint constructs shell commands by interpolating user-supplied gitName and gitEmail values into command strings passed to child_process.exec(). The input is placed within double quotes and only " is escaped, but backticks (`), $() command substitution, and \ sequences are all interpreted within double-quoted strings in bash. This allows authenticated attackers to execute arbitrary OS commands via the git configuration endpoint. This vulnerability is fixed in 1.24.0. |
| Craft is a content management system (CMS). The fix for CVE-2025-35939 in craftcms/cms introduced a strip_tags() call in src/web/User.php to sanitize return URLs before they are stored in the session. However, strip_tags() only removes HTML tags (angle brackets) -- it does not inspect or filter URL schemes. Payloads like javascript:alert(document.cookie) contain no HTML tags and pass through strip_tags() completely unmodified, enabling reflected XSS when the return URL is rendered in an href attribute. This vulnerability is fixed in 5.9.7 and 4.17.3. |
| Craft is a content management system (CMS). The ElementSearchController::actionSearch() endpoint is missing the unset() protection that was added to ElementIndexesController in CVE-2026-25495. The exact same SQL injection vulnerability (including criteria[orderBy], the original advisory vector) works on this controller because the fix was never applied to it. Any authenticated control panel user (no admin required) can inject arbitrary SQL via criteria[where], criteria[orderBy], or other query properties, and extract the full database contents via boolean-based blind injection. Users should update to the patched 5.9.9 release to mitigate the issue. |
| Craft is a content management system (CMS). Prior to 5.9.9 and 4.17.4, a Remote Code Execution vulnerability exists in the Craft CMS 5 conditions system. The BaseElementSelectConditionRule::getElementIds() method passes user-controlled string input through renderObjectTemplate() -- an unsandboxed Twig rendering function with escaping disabled. Any authenticated Control Panel user (including non-admin roles such as Author or Editor) can achieve full RCE by sending a crafted condition rule via standard element listing endpoints. This vulnerability requires no admin privileges, no special permissions beyond basic control panel access, and bypasses all production hardening settings (allowAdminChanges: false, devMode: false, enableTwigSandbox: true). Users should update to the patched 5.9.9 or 4.17.4 release to mitigate the issue. |
| Cursor is a code editor built for programming with AI. Prior to 2.0 ,if a visited website contains maliciously crafted instructions, the model may attempt to follow them in order to “assist” the user. When combined with a bypass of the command whitelist mechanism, such indirect prompt injections could result in commands being executed automatically, without the user’s explicit intent, thereby posing a significant security risk. This vulnerability is fixed in 2.0. |
| Striae is a firearms examiner's comparison companion. A high-severity integrity bypass vulnerability existed in Striae's digital confirmation workflow prior to v3.0.0. Hash-only validation trusted manifest hash fields that could be modified together with package content, allowing tampered confirmation packages to pass integrity checks. This vulnerability is fixed in 3.0.0. |
| Supabase Auth is a JWT based API for managing users and issuing JWT tokens. Prior to 2.185.0, a vulnerability has been identified that allows an attacker to issue sessions for arbitrary users using specially crafted ID tokens when the Apple or Azure providers are enabled. The attacker issues a valid, asymmetrically signed ID token from their issuer for each victim email address, which then is sent to the Supabase Auth token endpoint using the ID token flow. If the ID token is OIDC compliant, the Auth server would validate it against the attacker-controlled issuer and link the existing OIDC identity (Apple or Azure) of the victim to an additional OIDC identity based on the ID token contents. The Auth server would then issue a valid user session (access and refresh tokens) at the AAL1 level to the attacker. This vulnerability is fixed in 2.185.0. |
