CVE Feed

Use after free in CameraCapture in Google Chrome on Mac prior to 150.0.7871.128 allowed a remote attacker to potentially perform a sandbox escape via a crafted HTML page. (Chromium security severity: Critical)

AgenticMail gives AI agents real email addresses and phone numbers. In @agenticmail/claudecode prior to version 0.2.39, @agenticmail/codex prior to version 0.1.33, @agenticmail/core prior to version 0.9.43, and @agenticmail/openclaw prior to version 0.5.71, two inbound-mail handlers act on a privileged effect without verifying that the sender is the operator, while a sibling handler in the same repo does. The higher-impact one: any external email routed to the bridge inbox causes the dispatcher to resume the operator's Claude Code session with `permissionMode: 'bypassPermissions'`, embedding the attacker-controlled `from`/`subject`/`preview` verbatim into the prompt the resumed agent reads — an indirect prompt injection into a fully-privileged agent (Bash/Write/Edit/WebFetch + the agenticmail MCP toolbelt) running as the operator's OAuth identity. The sibling operator-query email-reply hook gates the same untrusted-From provenance with `isOperatorReplySender(replyFrom, config.operatorEmail)` (fail-closed); the bridge-wake path — a strictly higher-privilege effect — has no equivalent. @agenticmail/claudecode 0.2.39, @agenticmail/codex 0.1.33, @agenticmail/core 0.9.43, and @agenticmail/openclaw 0.5.71 contain a fix.

Grav CMS scheduler-webhook plugin contains an authentication bypass vulnerability that allows unauthenticated remote attackers to trigger configured scheduled jobs by exploiting a short-circuit logic flaw in the webhook token validation. Attackers can send a single unauthenticated POST request to the scheduler webhook endpoint to execute all configured scheduled jobs or target a specific job, causing unintended execution of operator-defined commands under the web server process user.

AgenticMail gives AI agents real email addresses and phone numbers. In @agenticmail/api prior to version 0.9.64, a low-privileged authenticated AgenticMail agent can enumerate another agent's pending/claimed tasks by supplying the target agent name to `GET /api/agenticmail/tasks/pending?assignee=<name>`. The returned task objects include the task IDs and payloads. The same task IDs can then be used with the capability-style task mutation endpoints (`/tasks/:id/claim`, `/tasks/:id/result`, `/tasks/:id/complete`, `/tasks/:id/fail`) to claim, complete, or fail tasks assigned to a different agent. Because ordinary authenticated agents can discover agent names through `GET /api/agenticmail/accounts/directory`, the task ID effectively stops being a secret capability. This turns the intended capability model into a cross-agent authorization bypass. Version 0.9.64 contains a fix.

AgenticMail gives AI agents real email addresses and phone numbers. @agenticmail/api prior to version 0.9.32 and @agenticmail/core prior to version 0.9.10 had weakness related to validation and and binding of inactive-agent hour filtering; storage SQL identifier validation; metadata-backed ownership checks for raw storage SQL; blocking direct storage metadata access through raw SQL; fail-closed outbound worker secret handling; SMTP envelope/header control-character validation before command construction; and TLS certificate verification as the default for MailSender with an explicit opt-out for local development. @agenticmail/api prior to version 0.9.32 and @agenticmail/core prior to version 0.9.10 are patched.

AVideo versions from commit 0dbadbca through latest master contain a server-side request forgery vulnerability in the encoder download-by-URL flow due to an unpinned retry fallback that bypasses DNS pinning validation. An authenticated attacker can supply a downloadURL that redirects to an internal address, causing the unpinned retry to follow the redirect and reach internal targets for blind SSRF attacks.

AVideo before 29.0 contains an incomplete fix for CVE-2026-45578 where execAsync() re-wraps escaped commands in double-quoted sh -c, allowing command substitution via $() and backticks. Attackers can inject arbitrary OS commands through the Live plugin on_publish.php endpoint despite escapeshellarg() protection.

FreeRDP before 3.28.0 treats lines beginning with forward slash in RDP files as raw command-line options, exposing the entire CLI parser surface to untrusted files. Attackers can craft malicious RDP files with /rdp2tcp, /cert:ignore, or /drive options to execute arbitrary commands, bypass certificate validation, or expose local filesystems without user interaction.

Shamefile is a linter for undocumented linter warnings. Prior to version 0.1.7, a path traversal vulnerability in `shame next` allows an attacker-controlled `shamefile.yaml` to disclose contents of files outside the repository, one line at a time, to the terminal of a user who runs the command. See patch commit for technical details. The issue is fixed in 0.1.7. Upgrade to either 0.1.7 or later versions to incorporate the patch. As a workaround, do not run `shame next` against untrusted `shamefile.yaml`. Use `shame me --dry-run` for CI validation.

nono is software that allows users to run AI agents in a zero-latency sandbox. Prior to version 0.55.0, the nono Landlock/seccomp policies allow access to local Unix domain sockets (concrete and abstract). This allows an easy sandbox escape by talking to the per-user systemd dbus socket. Version 0.55.0 patches the issue.

A vulnerability was identified in Metasoft 美特软件 MetaCRM up to 6.4.0 Beta06. The impacted element is an unknown function of the file /business/qnaire/upload.jsp. Such manipulation of the argument File leads to unrestricted upload. The attack may be launched remotely. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure but did not respond in any way.

The Spectra Gutenberg Blocks – Website Builder for the Block Editor plugin for WordPress is vulnerable to Stored Cross-Site Scripting via the `uagb/image` block in all versions up to, and including, 2.19.28 due to insufficient input sanitization and output escaping. This makes it possible for authenticated attackers, with Contributor-level access and above, to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page.

ClearanceKit intercepts file-system access events on macOS and enforces per-process access policies. The ECDSA private key used to sign the on-disk policy database (`/Library/Application Support/clearancekit/store.db`) is stored in the macOS System Keychain. The key was created via the two-step pattern `SecKeyCreateRandomKey` (in-memory) followed by `SecItemAdd(kSecValueRef:, kSecAttrAccess:)` (persist). Prior to version 5.0.10, for `kSecClassKey` items in the legacy System Keychain, `kSecAttrAccess` passed to `SecItemAdd` is silently ignored — the persisted key inherits no ACL restriction. The same access builder applied to `kSecClassGenericPassword` items correctly binds the ACL, making this bug specific to the EC key. The result is that any process running as root can use the key to produce valid signatures over arbitrary policy content. Version 5.0.10 fixes the issue. No known workarounds are available. Disabling the system extension and manually removing the System Keychain item labelled `clearancekit policy signing key` would prevent the forged-signature path but also disables policy enforcement.

ClearanceKit intercepts file-system access events on macOS and enforces per-process access policies. Prior to version 5.0.10, each table in the on-disk SQLite policy store (`/Library/Application Support/clearancekit/store.db`) is verified using an ECDSA signature stored in the `data_signatures` table. The signed payload contains only the canonical row content, with no version counter or freshness binding. An attacker who can write `store.db` and the matching `data_signatures` row — feasible during the opfilter-update window when the Endpoint Security filter is offline, or via offline-boot / decrypted-backup scenarios — can substitute a previously-captured legitimately-signed snapshot. opfilter accepts the older snapshot as fully valid on next boot because the existing signatures still verify. Version 5.0.10 patches the issue.

** REJECT ** DO NOT USE THIS CANDIDATE NUMBER. ConsultIDs: CVE-2026-43620. Reason: This candidate is a duplicate of CVE-2026-43620. Notes: All CVE users should reference CVE-2026-43620 instead of this candidate.

NextCRM is open-source customer relationship management (CRM) software. The CRM product catalog is an organization-wide business object. Normal application server actions restrict product creation, update, and deletion to `manager` and `admin` roles. However, in version 0.12.1, the MCP product tools expose the same write operations through `/api/mcp/mcp` using user-generated Bearer tokens and do not enforce role checks. Any authenticated low-privileged user who can generate an MCP API token can create, modify, archive, or soft-delete products in the shared CRM product catalog. Version 0.12.3 contains a fix.

NextCRM is open-source customer relationship management (CRM) software. In version 0.12.1, the MCP campaign tools expose campaign read and write operations over the network using user-generated Bearer API tokens (`nxtc__...`). The application has an authorization model that restricts normal users to campaigns they created, but multiple MCP campaign handlers ignore the authenticated user ID and query or mutate campaigns only by object ID. As a result, a low-privileged authenticated user with a valid MCP API token can enumerate all campaigns, read campaign details, update or delete campaigns owned by other users, modify campaign templates and steps, and potentially trigger or pause campaign delivery. Version 0.12.2 fixes the issue.

NextCRM is open-source customer relationship management (CRM) software. Versions prior to 0.12.0 have a Broken Object Level Authorization (BOLA/IDOR) vulnerability exists in the CRM contact and target update endpoints. The application fails to verify if the authenticated user has ownership of the specific resource being modified. This allows any authenticated user (even with a standard `member` role) to arbitrarily modify sensitive CRM contacts and targets belonging to other users or organizations (cross-tenant data tampering). Version 0.12.0 fixes the issue.

NextCRM is open-source customer relationship management (CRM) software. Versions prior to 0.12.0 have a Broken Access Control (BAC) vulnerability in the `activateUser` and `deactivateUser` Next.js Server Actions of NextCRM. The application fails to verify if the requesting user holds the `admin` role. Consequently, any authenticated user (even those with the lowest `member` or `viewer` roles) can arbitrarily activate or deactivate any user account in the system, including the main administrator. Version 0.12.0 fixes the issue.

** REJECT ** DO NOT USE THIS CANDIDATE NUMBER. ConsultIDs: CVE-2026-43618. Reason: This candidate is a duplicate of CVE-2026-43618. Notes: All CVE users should reference CVE-2026-43618 instead of this candidate.

** REJECT ** DO NOT USE THIS CANDIDATE NUMBER. ConsultIDs: CVE-2026-43617. Reason: This candidate is a duplicate of CVE-2026-43617. Notes: All CVE users should reference CVE-2026-43617 instead of this candidate.

Path traversal on Windows in Apache MINA SSHD component sshd-git. Apache MINA SSHD is a Java library for client-side and server-side SSH. A git server implemented with Apache MINA SSHD component sshd-git and running on Windows could allow an authenticated remote user access to git repositories outside of the configured server-side root directory. The path validation applied for CVE-2026-48827 in Apache MINA SSHD 2.18.0 and 3.0.0-M4 was partly ineffective for Servers running on Windows. Applications are affected if they use org.apache.sshd:sshd-git to implement a git server and run on Windows. Applications not using sshd-git or not running on Windows are not affected. Users are advised to upgrade affected applications to Apache MINA SSHD 2.19.0, which fixes the issue. The issue also is present in the pre-release milestones 3.0.0-M1 to 3.0.0-M4 for a new upcoming new major version 3.0.0. Again, applications are affected only if they use sshd-git and run on Windows. Upgrade affected applications to 3.0.0-M5.

Improper certificate validation in Apache MINA SSHD (server-side). Apache MINA SSHD is a Java library for client-side and server-side SSH. Server-side OpenSSH user certificate validation during user authentication in an Apache MINA SSHD server did not check for the unsupported force-command or verify-required options that could be embedded in the certificate, nor did it validate these options. As a result it was possible that a user could authenticate with such a certificate that included a force-command option but still was able to execute other commands. What other command exactly would be available to the user depends on the implementation of the server. This issue is fixed in Apache MINA SSHD 2.19.0 and 3.0.0-M5. Applications are advised to upgrade to these versions. The fix rejects OpenSSH user certificates that include these options, since Apache MINA SSHD implements neither force-command nor sk-*[email protected] user certificates (which are the only ones for which verify-required would make sense).

The `@ai-sdk/harness-opencode` tool is an HarnessV1 adapter backed by @openai/codex-sdk, which drives the codex command line interface. Prior to version 1.0.29, the tool relay authorizes requests from any process whose command line contains an allowed helper script path (the Codex CLI shim). This allows untrusted code executing in the sandbox to invoke arbitrary host-exposed tools, including secret lookups, deployment operations, and cloud API calls without a corresponding model-authorized tool-call event. Exploitation requires a Linux environment (the vulnerable fallback checks `process.platform === 'linux'` and reads `/proc`); an active harness session with one or more host-provided tools; and untrusted code executing in the sandbox (e.g. a malicious dependency, build script, or lifecycle hook) The fix in version 1.0.29 removes the process-path authorization fallback entirely. Relay requests are now only accepted after exact, short-lived, one-time authorization matching the tool name and input from a bridge-observed model event. Some workarounds are available. Do not run the Codex harness on untrusted repositories or with untrusted dependencies, and/or limit host-exposed tools to non-sensitive operations when working with untrusted code.

Improper input validation in sshd-git in Apache MINA SSHD. Apache MINA SSHD is a Java library for client-side and server-side SSH. Component org.apache.sshd:sshd-git provides though its GitPgmCommandFactory a way to configure an Apache MINA SSHD server such that SSH clients can remotely execute git commands via the JGit library on git repositories stored on the server. This GitPgmCommandFactory allowed a user authenticated via SSH to run any JGit command available, including commands that could write files at arbitrary places such as git archive with the --output option. Affected are SSH servers implemented with Apache MINA SSHD and using the GitPgmCommandFactory. If the GitPgmCommandFactory is not configured on the server, the server is not affected. It is recommended to upgrade affected servers to Apache MINA SSHD 2.19.0 or 3.0.0-M5, which fix this issue. The issue is fixed by restricting the available commands to a small whitelist of uncritical commands (such as git log). git archive is also allowed, but its --output argument is ignored and the archive is always sent through the SSH channel to the client.

** REJECT ** DO NOT USE THIS CANDIDATE NUMBER. ConsultIDs: CVE-2026-43619. Reason: This candidate is a duplicate of CVE-2026-43619. Notes: All CVE users should reference CVE-2026-43619 instead of this candidate.

Path traversal in the sshd-scp component of Apache MINA SSHD. Apache MINA SSHD is a Java library for client-side and server-side SSH. The implementation of receiving files or directories via SCP did not validate filenames in SCP "C" or "D" commands. A malicious sender could send filenames containing paths, resulting in files to be written in attacker-controlled places. The issue affects only * applications that use no longer supported Apache MINA SSHD versions < 2.0.0 and use the SCP functions to receive files, * or applications using sshd-scp in Apache MINA SSHD >= 2.0.0 to receive files. Applications using Apache MINA SSHD >= 2.0.0 not using sshd-scp are not affected. The issue is fixed in Apache MINA 2.19.0 and 3.0.0-M5. Affected applications are advised to upgrade to these versions.

The `@ai-sdk/harness-opencode` tool connects HarnessAgent to OpenCode through a sandboxed bridge. Prior to version 1.0.28, the tool relay authorizes requests from any process whose command line contains an allowed helper script path (`host-tool-mcp.mjs`). This allows untrusted code executing in the sandbox to invoke arbitrary host-exposed tools including secret lookups, deployment operations, and cloud API calls without a corresponding model-authorized tool-call event. Exploitation requires a Linux environment (the vulnerable fallback checks `process.platform === 'linux'` and reads `/proc`); an active harness session with one or more host-provided tools; and untrusted code executing in the sandbox (e.g. a malicious dependency, build script, or lifecycle hook) The fix in version 1.0.28 removes the process-path authorization fallback entirely. Relay requests are now only accepted after exact, short-lived, one-time authorization matching the tool name and input from a bridge-observed model event. Some workarounds are available. Do not run the OpenCode harness on untrusted repositories or with untrusted dependencies, and/or limit host-exposed tools to non-sensitive operations when working with untrusted code.

Paymenter is a free and open-source webshop solution for management of hosting services. In versions prior to 1.5.5, the credit payment implementation in app/Livewire/Invoices/Show.php executes a pessimistic row lock (lockForUpdate()) outside of an active database transaction. Because MySQL/MariaDB requires an enclosing transaction to enforce row-level locks, the guard is ineffective. Concurrent payment requests can exploit this race condition to read the same credit balance simultaneously, allowing users to pay multiple invoices using the same credit balance. In database systems like MySQL, a row lock only works inside a formal transaction; without one, the lock is completely ignored. Because there is no active lock, two payment requests sent at the exact same millisecond can look at the database at the same time. Both requests see the original credit balance, decide it is sufficient, and approve the payment. Because the payment processes successfully through ExtensionHelper::addPayment(), the application provisions the corresponding services or digital goods, resulting in direct financial or resource loss to the platform. This issue has been fixed in version 1.5.5.

Paymenter is a free and open-source webshop solution for management of hosting services. In versions prior to 1.5.1, the checkout component improperly filters URL-writable properties, allowing authenticated users to inject arbitrary key-value pairs into server provisioning parameters. Because bundled server extensions prioritize these user-supplied properties over administrator-defined configurations, a regular user can override hosting plans and resource limits at checkout without special privileges. The Checkout Livewire component's $checkoutConfig property exposed via URL query parameters, only validating keys explicitly defined by an extension's configuration method, allowing any undefined injected keys to bypass validation entirely. These unsanitized keys are then stored directly in the database by the cart component and later passed to server extensions during provisioning, enabling user-injected data to override intended administrator settings. Depending on the active extension, this leads to unauthorized overrides of core resource limits (such as CPU, RAM, storage, or package tiers). No administrative privileges are required to exploit this vulnerability. This issue has been fixed in version 1.5.1.

FreeScout is a free help desk and shared inbox built with PHP's Laravel framework. Prior to version 1.8.224, the FreeScout helpdesk application does not enforce rate limiting on the file upload endpoint. Any user can flood the server with upload requests, leading to database overload and potential denial of service for all users. Version 1.8.224 contains a fix.

FreeScout is a free help desk and shared inbox built with PHP's Laravel framework. Prior to version 1.8.224, the public endpoint `POST /user-setup/{hash}/{invite_sent_at}` (`OpenController@userSetupSave`) selects the target account solely by its `invite_hash` column, then overwrites that account's email and password and logs in as it. No authentication, cookie, or prior session is required. After a user activates, FreeScout sets `invite_hash` to the empty string. On MySQL and MariaDB, `VARCHAR` equality ignores trailing spaces, so a single URL-encoded space (`%20`) matches the stored empty string and selects the lowest-id activated user. The expiry guard decrypts `invite_sent_at` with the target's password hash, but `Helper::decrypt` returns its raw input unchanged when decryption fails. A plaintext numeric value such as `9999999999` therefore passes the time-to-live check without any secret. The result is that an anonymous attacker sets the email and password of the lowest-id activated FreeScout account (a support agent, or an administrator if one was added by invitation) and authenticates as that account. Version 1.8.224 contains a fix.

VSee Clinic 7.1.26 and API 1.3.0 contain an Insecure Direct Object Reference (IDOR) vulnerability in the /v1.3.0/api/files endpoint. An authenticated attacker can manipulate the 'remark' request parameter to enumerate, retrieve, and delete files belonging to other users on the application server.

FreeScout is a free help desk and shared inbox built with PHP's Laravel framework. FreeScout's `Manage -> Logs -> App Logs` feature uses the bundled `rap2hpoutre/laravel-log-viewer` override to decrypt a user-supplied file identifier and then pass the resolved path to Laravel's download response. Prior to version 1.8.224, the path resolution logic accepts any existing absolute path before applying the intended `storage/logs` restriction. As a result, an attacker who can access the App Logs route and forge a valid Laravel-encrypted `dl` parameter can download arbitrary server-local files readable by the PHP process, not just log files. Version 1.8.224 contains a fix.

VSee Clinic 7.1.26 and VSee Clinic API 1.3.0 exposes cleartext SFTP credentials in the HTTP responses of three unauthenticated endpoints. The credentials are present in these responses only when SFTP connections have been configured within the application. No authentication is required to retrieve these credentials. An unauthenticated remote attacker who observes any of these HTTP responses on an instance where SFTP is configured can obtain the credentials and use them to access the associated SFTP server.

Paymenter is a free and open-source webshop solution for management of hosting services. In versions prior to 1.5.0, the ticket creation endpoint accepts a user-supplied service identifier without enforcing ownership validation, allowing authenticated users to create support tickets referencing services belonging to other accounts by modifying the service ID in the request. An attacker could modify the service ID value in the client-side request and successfully create a ticket associated with another user's service. The vulnerability requires authentication and does not provide direct access to service contents or customer data. However, referenced service information could become visible to support personnel handling the ticket. Successful exploitation could allow an authenticated user to: create support tickets referencing services belonging to other users, potentially cause support staff to interact with or review unrelated customer services. The vulnerability did not allow direct access to another user's service, modification of another user's service or retrieval of confidential service data through the vulnerable endpoint itself. This issue has been fixed in version 1.5.0.

Paymenter is a free and open-source webshop solution for management of hosting services. In versions prior to 1.5.0, the PayPal webhook endpoint /extensions/paypal/webhook processes the PAYPAL-CERT-URL HTTP header without validation, allowing attackers to control server-side HTTP request destinations. This value is passed directly into a server-side HTTP request via file_get_contents, allowing attackers to control the destination of the request. No allowlist, validation, or signature verification is applied to the header before usage. As a result, the application can be coerced into performing HTTP requests to attacker-controlled or internal network destinations. This vulnerability allows remote unauthenticated attackers to induce server-side HTTP GET requests to arbitrary external or internal endpoints. Depending on network configuration, this may lead to: blind SSRF to external attacker-controlled systems, and potential access to internal network services No direct response data is returned to the attacker (blind SSRF), but the issue may still enable sensitive network probing or data exfiltration via side channels. This issue has been fixed in version 1.5.0.

Paymenter is a free and open-source webshop solution for management of hosting services. In versions prior to 1.5.0, the email update functionality fails to invalidate the existing verification state when a user changes their email address, allowing a verified account to retain its verified status after switching to an unverified or unowned email address. When a user updated their email address, the system did not reset or revalidate the associated email verification status. As a result, the verification column remained set to “true” even after the email address was changed. Exploitation could potentially result in: misrepresentation of email ownership, bypass of verification-based trust assumptions, and abuse of features gated behind verified status. No direct unauthorized access to other users accounts or data is possible through this issue alone. This issue has been fixed in version 1.5.0.

FreeScout is a free help desk and shared inbox built with PHP's Laravel framework. Prior to version 1.8.224, the denylist that neutralizes dangerous file uploads (`Helper::$restricted_extensions`) is incomplete: it does not cover the `.pht` extension. The authenticated upload endpoint `POST /uploads/upload` (`SecureController@upload`) stores files with their original extension into the web-accessible directory `storage/app/public/uploads/` (served at `/storage/uploads/`). On the standard Apache + `libapache2-mod-php` deployment, the default handler `<FilesMatch ".+\.ph(ar|p[3457]?|t|tml)$">` executes `.pht`, so **any authenticated agent can upload a `.pht` web shell and run arbitrary commands as the web-server user** (`www-data`). This is a direct bypass of the fix for CVE-2025-48471, which added `phtml`/`phar` but not `pht` (nor `phtm`, `phps`). Version 1.8.224 contains an updated fix.

FreeScout is a free help desk and shared inbox built with PHP's Laravel framework. A Prototype Pollution condition in the `getQueryParam` function `/public/js/main.js` and was addressed in version 1.8.139 by blocking URL query keys matching the pattern `__proto__`. However, this mitigation is incomplete: it only filters top-level `__proto__` keys and fails to sanitize nested forms such as `b[__proto__][polluted]=PWNED`. As a result, an attacker-controlled URL query string can still write into `Object.prototype` on any page that loads `main.js`. Version 1.8.223 contains a updated fix.

FreeScout is a free help desk and shared inbox built with PHP's Laravel framework. Prior to version 1.8.223, an unauthenticated attacker can inject messages into any existing support conversation by sending a single email to the helpdesk's public address with a crafted `In-Reply-To` header. No credentials, tokens, or prior access are required. The injected message is rendered in the agent UI as a legitimate customer reply, the conversation is automatically reopened, and the `last_reply_from` field is set to the attacker's identity. Version 1.8.223 contains a fix.

This CVE ID has been rejected or withdrawn by its CVE Numbering Authority.

RT is an open source, enterprise-grade issue and ticket tracking system. Versions 5.0.4 up to (but not including) 5.0.10, and 6.0.0 up to (but not including) 6.0.3 contain a reflected Cross-Site Scripting (XSS) vulnerability where an attacker who can induce an authenticated RT user to visit a crafted URL can execute arbitrary JavaScript in that user's browser session. This issue has been fixed in versions 5.0.10 and 6.0.3.

RT is an open source, enterprise-grade issue and ticket tracking system. Versions prior to 5.0.10, 6.0.0 and above, prior to 6.0.3 contain an information disclosure and privilege escalation vulnerability in the REST 2.0 API. A privileged (non-administrative) user can obtain authentication credentials belonging to other users — including users with administrative privileges — and use those credentials to read data as those users via RT's feed endpoints. The same request that exposes the credentials also rotates them, invalidating previously-distributed feed URLs across the instance. This issue has been fixed in versions 5.0.10 and 6.0.3.

RT is an open source, enterprise-grade issue and ticket tracking system. Versions 5.0.0 and 6.0.0 and above, prior to both 5.0.10 and 6.0.3 contain a Cross-Site Scripting (XSS) vulnerability where uploaded content is served inline rather than as an attachment. An authenticated user who can upload content can include JavaScript in the upload that will execute in the browser session of any RT user who later views or downloads it. This issue has been fixed in versions 5.0.10 and 6.0.3.

GPT-SoVITS through 20250606v2pro contains an OS command injection vulnerability in webui.py where ASR, slice, denoise, and uvr5 functions interpolate unsanitized Gradio textbox values directly into shell commands executed with shell=True. Attackers can inject shell metacharacters through path parameters to execute arbitrary OS commands as the server process user without authentication.

Improper authorization in the ToolGroupResource and RoleAjax REST/DWR endpoints in dotCMS dotCMS 21.02 through 26.06.22-03 on all platforms allows a low-privileged authenticated backend user to self-assign the administrative layout and self-grant the CMS Administrator role, then achieve remote code execution via a crafted OSGi bundle upload whose BundleActivator executes arbitrary shell commands.

ktransformers through 0.6.3, fixed in commit def0f93, contains an unauthenticated pickle deserialization vulnerability that allows remote attackers to execute arbitrary commands by sending crafted pickle payloads to the SchedulerServer ZMQ ROUTER socket bound to all interfaces. Attackers can exploit malicious __reduce__ methods embedded in crafted pickle payloads to execute arbitrary shell commands as the server process.

BuildKit's cache mount source= selector on Windows Container on Windows (WCOW) workers does not detect NTFS directory junctions placed inside the cache root. A build authored by an untrusted user on a WCOW-configured BuildKit daemon can read arbitrary host files reachable to the BuildKit daemon process.

cal.diy through 6.2.0 contains an open redirect vulnerability in the conferencing OAuth callback endpoint that allows attackers to redirect users to arbitrary URLs by crafting malicious state parameters. Attackers can exploit the unsigned state parameter and onErrorReturnTo field to silently redirect visitors from the trusted domain to attacker-controlled URLs for phishing attacks.