🚨 CVE-2026-54234
vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Prior to 0.24.0, a frontend-legal multi-request speculative decoding workload can cause the rejection sampler to produce a recovered token equal to the model vocabulary size boundary value, which is then converted to negative one when the engine selects the next live token for a request and is written back into the drafter's input ids; that out-of-vocabulary value is later consumed by the model's embedding and attention path and crashes the engine worker with a GPU device-side assertion. The same triggering request sequence is reachable through the public gRPC Generate and Abort endpoints, so a remote client that can send generation requests can crash the shared engine worker, aborting concurrent requests and causing a service-wide denial of service for other clients of the deployment until the worker is restarted. This issue is fixed in version 0.24.0.

🎖@cveNotify

🚨 CVE-2026-55514
vLLM is a library for LLM inference and serving. From 0.12.0 to before 0.24.0, sending a pure prompt embeds payload in a /v1/completions request with a model using M-RoPE causes EngineCore to fail an assertion and fatally crash, shutting down the entire server application. Any remote user who is authorized to make a /v1/completions request can make such a request and induce a crash. This issue is fixed in version 0.24.0.

🎖@cveNotify

🚨 CVE-2026-55574
vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Prior to 0.24.0, the structured_outputs.regex API parameter passes a user-supplied regular expression string directly to the grammar compiler backends with no compilation timeout; in the xgrammar backend the string reaches the regex compiler with no guard, and in the outlines backend the validation step blocks structural issues such as lookarounds and backreferences but performs no complexity analysis, so a pattern with nested quantifiers passes all checks and causes exponential state-space expansion, allowing a single request containing an adversarial regex to hang an inference worker indefinitely and deny service. This issue is fixed in version 0.24.0.

🎖@cveNotify

🚨 CVE-2026-57571
Crawl4AI is an open-source LLM-friendly web crawler and scraper. Prior to 0.9.0, when the crawler saves a downloaded file, the destination filename was taken from attacker-influenced input and joined to the downloads directory with no confinement. A filename containing an absolute path or traversal escaped the downloads directory, giving an arbitrary file write with attacker-controlled contents; the HTTP crawler path uses the response Content-Disposition filename and the browser crawler path uses the download's suggested filename. Because the written bytes are attacker-controlled, this can escalate to remote code execution. This issue is fixed in version 0.9.0.

🎖@cveNotify

🚨 CVE-2026-57572
Crawl4AI is an open-source LLM-friendly web crawler and scraper. Prior to 0.9.0, the Docker API server accepted request-supplied browser_config.extra_args, which flowed into Chromium's launch arguments. An attacker could inject Chromium switches that replace a child-process launch command together with --no-zygote, causing Chromium to fork or exec an attacker-controlled command as the container's runtime user. The Docker API is unauthenticated by default, so a single request yields arbitrary command execution. This issue is fixed in version 0.9.0.

🎖@cveNotify

🚨 CVE-2026-57573
Crawl4AI is an open-source LLM-friendly web crawler and scraper. Prior to 0.9.0, the Docker API server applied its SSRF destination check on the non-streaming /crawl path but not on the streaming path. handle_stream_crawl_request passed seed URLs straight to the crawler with no destination validation, allowing a remote unauthenticated client to call POST /crawl/stream or POST /crawl with crawler_config.stream=true with a URL pointing at an internal, private, or link-local address; the server fetched it and streamed the response body back. This issue is fixed in version 0.9.0.

🎖@cveNotify

🚨 CVE-2026-11610
A heap buffer overflow flaw was found in the SASL I/O layer of 389 Directory Server
(389-ds-base). After a successful SASL bind with integrity protection (SSF > 0),
an authenticated attacker can send a specially crafted oversized LDAP UNBIND packet
that is copied into a 512-byte heap receive buffer without a bounds check in
sasl_io_recv() in sasl_io.c. This allows up to approximately 2 megabytes of
attacker-controlled data to overflow the buffer, causing a denial of service (server
crash). In FreeIPA and Red Hat Identity Management deployments, any domain user with
a valid Kerberos ticket, any enrolled host, or any service account can trigger this
vulnerability over the network after authenticating via GSSAPI.
The vulnerable code path has existed since approximately 2013 (389-ds-base 1.3.2) and
was not addressed by the CVE-2025-14905 fix, which patched a separate heap overflow
in schema.c only.

🎖@cveNotify

🚨 CVE-2025-12799
A flaw was found in Jastow. Jastow is vulnerable to Cross-Site Scripting (XSS) attack. If using a set of combined configuration to allow unescaped characters in URL with embedded Undertow and Jastow, a server might be vulnerable to improper input handling.

🎖@cveNotify

🚨 CVE-2026-23697
Vtiger CRM before 8.4.0 contains an authenticated file upload vulnerability that allows low-privileged users to achieve remote code execution by uploading a .phar file containing arbitrary PHP code through the Documents module, bypassing the extension denylist in config.inc.php which omits the .phar extension. The uploaded file is stored with its original .phar extension under the web-accessible storage directory, and a misconfigured .htaccess using Apache 2.2 syntax is silently ignored on Apache 2.4 deployments, allowing unauthenticated HTTP requests to directly execute the uploaded PHP payload.

🎖@cveNotify

🚨 CVE-2026-48947
An improper access check allows privileged users to overwrite media files without editing permissions.

🎖@cveNotify

🚨 CVE-2026-48948
An improper access check allows user to download vcard exports of com_contact contacts that are inaccessible.

🎖@cveNotify

🚨 CVE-2026-48949
Lack of validation leads to an XSS vulnerability in the MFA management views.

🎖@cveNotify

🚨 CVE-2026-48951
Lack of escaping leads to XSS vulnerabilities in modalreturn layouts of various components.

🎖@cveNotify

🚨 CVE-2026-48952
Lack of escaping leads to an XSS vulnerability in the update list view of com_installer.

🎖@cveNotify

🚨 CVE-2026-48953
Lack of escaping leads to an XSS vulnerability in the generic image output layout.

🎖@cveNotify

🚨 CVE-2026-48954
Improper validation leads to a generic XSS vector in the language override feature.

🎖@cveNotify

🚨 CVE-2026-48955
An improper access check allows unauthorized users to access workflow stage and transition information.

🎖@cveNotify

🚨 CVE-2026-48956
An improper access check allows users to display a list of modules in the frontend.

🎖@cveNotify

🚨 CVE-2026-48957
An improper access check allows unauthorized users to access com_privacy datasets.

🎖@cveNotify

🚨 CVE-2026-48958
An improper access check allows unauthorized users to create custom fields via webservices endpoints.

🎖@cveNotify

🚨 CVE-2026-55435
Coder allows organizations to provision remote development environments via Terraform. Starting in version 2.30.0 and prior to versions 2.32.7, 2.33.8, and 2.34.2, AI Bridge proxy endpoints authenticate via `Server.IsAuthorized` in `coderd/aibridgedserver`, which validates key format, expiry, secret and deleted or system users but does not check whether the account is suspended. Because suspension does not revoke existing API keys, a suspended user's unexpired token keeps working. Practical impact is limited to already-issued API keys of suspended users until those keys are deleted. Versions 2.32.7, 2.33.8, and 2.34.2 patch the issue. As a workaround, on suspension, delete the user's API keys via `DELETE /api/v2/users/{user}/keys`.

🎖@cveNotify