CVE Feed

Privilege Escalation via Access Token Scope Escalation in API

Email Management API Bypasses ManageCredentials Feature Restrictions

Null Pointer Dereference in AddTime API Causes Authenticated Denial of Service

OIDC userinfo Endpoint Returns Identity Claims Without Enforcing API Token Scopes

Internal API HTTP client hardcodes InsecureSkipVerify:true with no config override (CWE-295)

RSS/Atom feed handlers bypass API-token scope & public-only confinement (incomplete fix of #37698)

Denial of Service via Unbounded io.ReadAll in NPM Package Tag Endpoint

Blind SSRF in OAuth2 avatar synchronization via unvalidated OIDC picture claim

SQL injection in the Zalktis accounting application via trading-partner-controlled text fields in received electronic invoices. When importing a received e-invoice (UBL/PEPPOL) or an e-commerce export, Zalktis concatenates partner-controlled values directly into SQL statement text using string concatenation, with neither parameterised queries nor escaping. The application's own escaping helper, Dazadi.sql_txt(), is not invoked on these code paths, so a party that sends an invoice can break out of the string literal and alter the query logic. This issue affects Zalktis: before 2026.1.586 and before 2026.2.592.

Form::Processor::Field::HtmlArea versions from 0.06 through 1.162360 for Perl allow attacker selected method dispatch and resource exhaustion via an HTML::Tidy diagnostic that validate passes to add_error as a Locale::Maketext template. validate runs HTML::Tidy over the submitted markup and passes each resulting message to add_error as its first argument, which add_error hands to the language handle as the Locale::Maketext message key. The default handle's lexicon sets `_AUTO`, so a message that is not a lexicon entry is compiled as a bracket notation template instead of being looked up. Tidy diagnostics quote the offending attribute name or value, so a bracket group in the submitted markup reaches the template position, where the first token of the group names a method called on the language handle and the remaining tokens are its arguments. A group such as `[0]` makes the compile croak, and neither the field nor the handle catches it, so the exception leaves validate. `[sprintf,%2000000000d,7]` reaches CORE::sprintf with an attacker chosen field width. One submission of crafted markup to an HtmlArea field throws an unhandled exception out of form validation or allocates an arbitrary amount of memory, and an application whose language handle subclass defines side effecting public methods makes those callable with attacker chosen arguments. The other field types pass fixed templates with the submitted value in an argument slot, where it stays inert, and are unaffected.

Data::MuForm::Localizer versions through 0.05 for Perl execute Perl from a message catalog header, reached at an arbitrary path because load_lexicon interpolates the language attribute into the catalog filename. load_lexicon builds the catalog path by appending `Messages/$lang.po` to the directory holding Localizer.pm, where $lang is the language attribute, with no check that it names a bare locale tag. A value holding `../` segments walks out of the message directory, so any readable path with a `.po` suffix is loaded. While parsing the catalog, extract_header_msgstr takes the `Plural-Forms:` header, prefixes `$` to the bare words nplurals, plural and n, and passes the rest verbatim into a string that is evaluated: the nplurals form evaluates the header expression immediately, and the plural_code form compiles it into a subroutine whose body runs when a plural message is localized. A header of `nplurals=2; plural=(system('...'),0);` therefore runs that command as the catalog loads. The evaluation inherits strict, so an expression that assigns to an undeclared variable fails to compile, while one built from calls alone does not. An application that sets the language attribute from request data, an Accept-Language header or a locale parameter, and an attacker who can place a file with a `.po` suffix and chosen contents at a readable path, together give code execution as the application user. The message expansion path is not affected: expand_named substitutes only the placeholder names the caller supplies, and _mangle_value returns the value unchanged.

HTML::FormHandler versions before 0.410000 for Perl allow attacker selected method dispatch and resource exhaustion because _apply_actions and add_error use error message text built from request data as a Locale::Maketext bracket notation template. add_error hands its first argument to the language handle as the Locale::Maketext message key, and the default handle's lexicon sets `_AUTO`, so a string that is not a lexicon entry is compiled as a bracket notation template instead of being looked up. In a bracket group the first token names a method called on the language handle and the remaining tokens are its arguments. Three kinds of text the library did not author reach that position. _apply_actions installs a `$SIG{__WARN__}` handler that stores the warning text in `$error_message`, and a captured warning survives a successful action, so a field carrying a numeric transform turns `Argument "[sprintf,%50000000d,0]" isn't numeric` into the template; a warning quotes the submitted value verbatim, so the group is well formed and dispatches. `$error_message ||= $tobj->validate($new_value)` takes a type constraint's own failure message, which renders the rejected value through a partial dumper in bracket and comma form (Devel::PartialDump when Moose can load it, Type::Tiny's own dumper always), so a field with `apply => [ Str ]` given a parameter sent more than once, which arrives as an array, gets `Reference ["a","b"] did not pass type constraint "Str"` as its template, from a request that carries no bracket character of its own. A coercion or transform exception reaches it the same way. Beyond those, a validator whose message contains the field value puts that value in the template directly, and add_error replaces the message list with the contents of an arrayref first argument (`@message = @{$message[0]} if ref $message[0] eq 'ARRAY'`), so a value arriving as an array fills the argument slots from the same request as well. A malformed group such as `[0]` makes the compile croak, and HTML::FormHandler::I18N::maketext and add_error each re-raise that as a die, so process() throws. A well formed group naming sprintf reaches CORE::sprintf with an attacker chosen field width. Any caller that applies a type constraint or a transform to an untrusted field, or whose validator passes an untrusted field value to add_error, can be made to throw an unhandled exception out of process(), or to allocate an arbitrary amount of memory in one request, and an application whose language handle subclass defines side effecting public methods makes those callable with attacker chosen arguments. The dumped type constraint message is bounded to the exception, because both dumpers quote non-numeric elements so the method slot is never an attacker chosen name. The built-in messages pass fixed templates with the value in an argument slot, where it stays inert, and the built-in field types attach explicit message callbacks, so neither is affected.

CVE-2026-55402 is an out of bounds read vulnerability in Secure Access servers prior to version 14.57. Attackers with an ‘in the middle’ position can send specially crafted data to a server causing a persistent denial of service.

CVE-2026-55401 is a null dereference vulnerability on the load-balancing sub-system of Secure Access servers prior to 14.57. Attackers can send an unauthenticated packet to a Secure Access server with load balancing enabled, which results in the internal load balancer crashing. After a successful attack, the Secure Access server is still able to accept connections and is still able to issue a failover to connected clients. ‍ https://www.first.org/cvss/calculator/4.0#CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:L

CVE-2026-55400 is an integer underflow in Secure Access servers prior to version 14.57. Attackers with an authenticated session can send specially crafted traffic to a server in a non-default configuration and cause a persistent denial of service.

Fluent Forms Pro 6.2.7 contains an embedded malicious code vulnerability introduced via a tampered plugin build served through a decommissioned update server. The tampered build introduced a rogue PHP file (libs/class-license-sync.php), loaded via a require_once directive added to fluentformpro.php, that established a backdoor REST API endpoint, dropped persistent PHP files in mu-plugins and uploads directories, installed a passwordless administrator account, and registered scheduled tasks that survived plugin removal.

Ninja Tables Pro 5.2.11 contains an embedded malicious code vulnerability introduced via a tampered plugin build served through a decommissioned update server. The tampered build introduced a rogue PHP file (app/Library/updater/NinjaTableDataSync.php) that established a backdoor REST API endpoint, dropped persistent PHP files in mu-plugins and uploads directories, installed a passwordless administrator account, and registered scheduled tasks that survived plugin removal.

Saurus CMS Community Edition contains an unauthenticated open redirect vulnerability in the logout handling code in classes/port.inc.php, where the url parameter supplied via GET or POST is passed directly to the Location header without domain allowlist, scheme validation, or relative path enforcement. Attackers can craft a malicious logout URL containing an arbitrary external domain or javascript: URI scheme to redirect authenticated users to attacker-controlled phishing pages after session destruction, enabling credential theft and OAuth redirect abuse.

Perl versions from 5.9.4 before 5.41.9 produce incorrect regular expression match results when a stale failure flag ends the Aho-Corasick prescan early in S_find_byclass. The prescan walks the subject for positions where the full pattern could match, and the engine tries it from the leftmost one recorded. A failing transition sets the failed flag, and a later successful transition does not clear it, so the prescan reads the stale flag as a failure and stops before it can record a candidate that starts earlier. It takes a subject where one candidate is recorded and a later character then forces a fallback through a fail link that succeeds. Example: "ABCDE" =~ m/ABCF|BCDE|C/; # matches C at offset 2, not BCDE "ABCDE" =~ m/ABCF|BCDE|C(G)/; # no match, BCDE missed An alternation like this can miss input it should match, or match it on the wrong branch, so an access or filtering decision made from the result can be wrong.

Cross-site Scripting in the Markdown renderer in maalfer Pentestify before 2.3.2 allows authenticated users to execute arbitrary JavaScript in the application origin via a Markdown link whose URL contains a double quote, which closes the anchor's href attribute because the renderer's sanitization step does not escape quotes

PostGIS before 3.7.0beta2 contains an out-of-bounds read vulnerability that allows attackers to cause memory disclosure or a server crash by supplying a malformed FlatGeobuf buffer. The FlatGeobuf property metadata decoder verifies that a string length field is present but fails to verify that the subsequent string body is contained within the supplied buffer before materializing it into a SQL-visible value, enabling memory disclosure or denial of service.

The address_standardizer extension for PostGIS through 3.7.0, fixed in commit 423570b, contains an out-of-bounds write vulnerability that allows a database user with the ability to supply caller-controlled relation names to standardize_address() to trigger memory corruption by providing a rules table with a classification Type value exceeding the fixed class range. Attackers can craft a malicious rules table entry with an oversized rule type value that is used without bounds checking as an index into an internal output-link table, resulting in an out-of-bounds write.

A CMS contains a SQL injection vulnerability in admin/db_data.php at line 509 that allows authenticated administrators to inject arbitrary SQL into a SHOW COLUMNS FROM statement by supplying unsanitized input through the table_name GET or POST parameter. Attackers can perform table traversal, time-based blind, boolean-based blind, and error-based injection techniques to enumerate full database schema, access system tables such as information_schema, and chain the disclosure with secondary injection points to extract credential data.

In Zimbra Collaboration (ZCS) before 10.1.17, weak cryptographic key generation vulnerability exists in the OnlyOffice integration. The zimbraDocumentEditingJwtSecret is generated using an insecure random number generator, resulting in insufficient entropy. An attacker who obtains a JWT signed with the generated secret may be able to recover the JWT signing secret through offline brute-force, potentially enabling JWT forgery.

In Zimbra Collaboration (ZCS) before 10.1.17, a Cross-Site Request Forgery (CSRF) vulnerability exists in the Exchange Web Services (EWS) endpoint of Zimbra Collaboration (ZCS) due to insufficient validation of request content types. An attacker can exploit this vulnerability by causing an authenticated user to submit a crafted request, potentially allowing unauthorized actions to be performed on behalf of the victim.

In Zimbra Collaboration before 10.1.17, a local file inclusion (LFI) vulnerability exists in the Zimbra Classic Web Client due to improper validation of the fu request parameter. An unauthenticated attacker can exploit this vulnerability by supplying a crafted path, potentially allowing unauthorized disclosure of protected files, such as WEB-INF/web.xml, within the web application directory. This occurs in the Forward servlet.

In Zimbra Collaboration (ZCS) before 10.1.17, a path traversal vulnerability exists in the Zimbra Briefcase document editing functionality due to improper validation of the packages parameter. An authenticated attacker can exploit this vulnerability by supplying a crafted path traversal sequence, potentially allowing unauthorized disclosure of sensitive files within the web application directory.

In Zimbra Collaboration (ZCS) before 10.1.17, a stored cross-site scripting (XSS) vulnerability exists in the Zimbra Classic Web Client due to insufficient sanitization of specific attachment content during inline preview. An attacker can send a crafted email containing a malicious attachment that, when previewed by a user, executes arbitrary JavaScript within the victim's browser session. Successful exploitation may allow an attacker to perform unauthorized actions on behalf of the victim user, potentially leading to data exfiltration or unauthorized access to sensitive information.

An authorization bypass vulnerability exists in Zimbra Collaboration (ZCS) before 10.1.17 due to improper authorization validation in delegated email sending functionality. An authenticated attacker can send specially crafted SOAP requests to impersonate another user and send emails without possessing the required delegation or send-as permissions. This occurs in the SaveDraftRequest SOAP handler.

A remote code execution vulnerability exists in Zimbra Collaboration (ZCS) before 10.1.20 when the optional zimbra-snmp package is installed and SNMP notifications are enabled. Due to improper sanitization of untrusted input during SNMP notification processing, an unauthenticated attacker can send specially crafted SMTP requests that may result in execution of arbitrary operating system commands as the Zimbra user.

A vulnerability was found in SourceCodester Simple Student Information System. Affected by this vulnerability is an unknown functionality of the file app/admin/departments/view_department.php. Performing a manipulation of the argument ID results in sql injection. The attack is possible to be carried out remotely. The exploit has been made public and could be used.

vLLM is an inference and serving engine for large language models. From 0.19.0 until 0.26.0, the /v1/completions CompletionRequest.prompt field in vllm/entrypoints/openai/completion/protocol.py accepts an unbounded list[str] or list[list[int]], prompt_to_seq() in vllm/renderers/inputs/preprocess.py and OnlineRenderer.preprocess_completion() in vllm/renderers/online_renderer.py expand every element, and vllm/entrypoints/openai/completion/serving.py creates one engine generator and response slot per prompt, allowing an authenticated API client to exhaust CPU, memory, async scheduling capacity, engine request slots, and response buffering with one request. This issue is fixed in version 0.26.0.

vLLM is an inference and serving engine for large language models. Prior to 0.27.0, an integer overflow in blockIdx.x * 2 * d in activation_kernels.cu can cause act_and_mul_kernel to consume another batched user's input, allowing a request processed in the same inference batch to receive a partial or complete copy of another user's inference result. This issue is fixed in version 0.27.0.

@jshookmcp/jshook is an MCP server that gives AI agents tools for JavaScript analysis and security research. In version 0.3.1, he network domain has a central SSRF authorization policy that blocks private, loopback, link-local, and reserved targets unless an explicit authorization object allows private network access. The policy is enforced by raw HTTP/TCP/TLS RTT tools, but the ICMP probe and traceroute tools resolve the target and invoke the native ICMP/traceroute sink directly. An MCP client with access to an active network domain can therefore ask the jshookmcp server to probe internal addresses even when local SSRF access is disabled for the other raw network tools. This exposes an internal reachability and route mapping primitive from the server network position. Version 0.3.2 fixes the issue.

vLLM is an inference and serving engine for large language models. From 0.20.2rc0 until 0.26.0, safe_load_prompt_embeds in vllm/renderers/embed_utils.py uses torch.sparse.check_sparse_tensor_invariants, whose process-global save, enable, and restore state can be raced by concurrent prompt_embeds parts submitted to POST /v1/chat/completions through AsyncMultiModalItemTracker.resolve_items, asyncio.gather, and the default executor, allowing an invalid sparse tensor to reach tensor.to_dense despite the CVE-2025-62164 guard when enable_prompt_embeds is enabled. This issue is fixed in version 0.26.0.

vLLM is an inference and serving engine for large language models. Prior to 0.26.0, the structured_outputs.regex parameter in vllm/v1/structured_output/backend_lm_format_enforcer.py is passed to lmformatenforcer.RegexParser without compile_regex_with_timeout or validation in validate_structured_output_request_lm_format_enforcer, allowing an unauthenticated /v1/completions request against the lm-format-enforcer backend to consume a CPU core and stall the structured-output engine path with a catastrophic regular expression. This issue is fixed in version 0.26.0.

ELAN reported a potential out-of-bounds write vulnerability in the ELAN TrackPoint driver that, under certain circumstances, could allow a local authenticated user to cause a system crash.

During an internal security assessment, an improper link following vulnerability was identified in Lenovo Vantage and Lenovo Commercial Vantage that could allow a local authenticated user to execute code with elevated privileges.

During an internal security assessment, a potential vulnerability was discovered in Lenovo Accessories and Display Manager for Enterprise for Windows that could allow a local authenticated user to execute arbitrary code with elevated privileges.

During an internal security assessment, an improperly protected key was discovered in Lenovo Dock Manager that could allow a local authenticated user to escalate privileges.

During an internal security assessment, a potential improper permissions vulnerability was discovered in Lenovo Dock Manager that could allow a local authenticated user to execute arbitrary code with elevated privileges.

During an internal security assessment, a potential vulnerability was discovered in Lenovo Dock Manager that could allow an authenticated local user to perform an arbitrary file deletion with elevated privileges.

An improper link following vulnerability was reported in the VantageCoreAddin for Lenovo Vantage and Lenovo Commercial Vantage that could allow a local authenticated user to perform an arbitrary file deletion with elevated privileges.

A potential authentication bypass vulnerability was reported in Lenovo System Update that could allow a local authenticated user to execute arbitrary code with elevated privileges.

auth-fetch-mcp is an MCP server that lets AI assistants fetch content from authenticated web pages. Version 3.0.1 implements SSRF protection in `assertSafeUrl()` (`src/security.ts`) to block requests to private and loopback addresses. However, the `isPrivateV6()` function fails to detect IPv4-mapped IPv6 loopback addresses in their hex-normalized form. When an attacker supplies a URL such as `http://[::ffff:127.0.0.1]:PORT/`, the Node.js WHATWG URL parser silently normalizes the host to `[::ffff:7f00:1]`. Because `net.isIPv4('7f00:1')` returns `false`, the private-IP check is bypassed and the URL is passed to the browser or HTTP client, allowing the MCP tool to reach loopback services that are supposed to be blocked. The issue is exploitable under default configuration without any special environment variable. Version 3.0.1 patches the issue.

vLLM is an inference and serving engine for large language models. Prior to 0.26.0, the validation_exception_handler in vllm/entrypoints/openai/server_utils.py converts FastAPI RequestValidationError objects with str(exc), and sanitize_message in vllm/entrypoints/utils.py does not remove traceback-style file paths, allowing unauthenticated malformed JSON requests to /v1/chat/completions, /v1/completions, /tokenize, and /detokenize to disclose the OS username, home and virtual-environment paths, Python version, internal package structure, line numbers, and endpoint handler names. This issue is fixed in version 0.26.0.

rsync 3.1.0 before 3.5.0 contains an access control bypass vulnerability that allows remote attackers to circumvent hosts deny rules by inducing DNS resolution failures during hostname-based access control evaluation. When a DNS lookup for a hostname-based deny rule fails, the daemon skips the rule rather than defaulting to a deny decision, enabling attackers who can trigger DNS failures to bypass module-level IP access controls and gain unauthorized access to restricted module file trees.

rsync before 3.5.0 contains an algorithmic complexity vulnerability in the hash_search() function that allows a remote attacker to cause a denial of service by delivering a carefully constructed file list. A sender can exploit the quadratic-time worst-case behavior in hash lookups to exhaust receiver CPU resources with a modest number of crafted entries, causing a sustained denial of service.

rsync 3.2.0 through 3.2.3 (openssl mode) and rsync-ssl through 3.4.4 (stunnel mode) contain a TLS certificate validation vulnerability that allows on-path attackers to intercept encrypted sessions by presenting self-signed or otherwise invalid certificates. Attackers can exploit the failure to validate server TLS certificates against a trusted CA or verify certificate hostname matching to decrypt or tamper with rsync session content without detection by the client.

rsync 3.4.2 before 3.5.0 contains a denial of service vulnerability that allows a remote sender to exhaust system resources by specifying the --zt short alias for --compress-threads, which bypasses the refuse options directive's string matching on long option names. Attackers can specify --zt=N with a large value to spawn an unbounded number of Zstandard worker threads on the receiver, exhausting available thread and memory resources.