| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton _bincount_kernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing out-of-bounds writes that corrupt concurrent requests' sampler state and alter repetition penalty behavior. |
| uri-js through 4.4.1 contains an improper UTF-8 decoding vulnerability in pctDecChars() that decodes invalid and overlong percent-encoded sequences into ASCII metacharacters. Attackers can craft percent-encoded payloads to bypass platform decoder validation and inject path traversal or CRLF sequences that downstream consumers process without filtering. |
| Azkaban through 4.0.0 omits project permission checks in the ScheduleServlet fetchSchedule action, allowing authenticated users to read any project's schedule configuration. Attackers can supply arbitrary project and flow identifiers to retrieve sensitive schedule details including execution times, cron expressions, flow parameters, and notification email lists without proper authorization. |
| SGLang through 0.5.19 in prefill/decode disaggregation mode with Mooncake KV transfer backend fails to validate bootstrap_room values, allowing unbounded transfer state allocation. Unauthenticated attackers can reach the decode engine's POST /generate endpoint and submit arbitrary bootstrap_room values to exhaust prefill process memory until out-of-memory termination. |
| A flaw was found in the multicluster-observability-addon. A remote attacker can access a debug endpoint without authentication, due to a misconfiguration in the underlying addon-framework library. This allows for the disclosure of sensitive operational information, such as goroutine, heap, and command-line details, after completing a basic encrypted connection. This vulnerability does not enable direct remote code execution. |
| hickory-resolver versions before 0.26.2 fail to propagate bogus DNSSEC proof states through the Resolver::lookup() and Resolver::lookup_ip() APIs, allowing invalid records to be returned as successful results. Attackers controlling the answering zone or positioned on the network path can have forged DNS records accepted as validated, bypassing DNSSEC authentication checks. |
| vm2 NodeVM versions before 3.12.1 contain a sandbox escape vulnerability where the DANGEROUS_BUILTINS denylist omits child_process despite blocking other host-spawning modules. Attackers can require child_process and execute arbitrary commands on the host system when NodeVM is configured with builtin:['*'] or explicit child_process allowance. |
| rustls-webpki (rustls/webpki) versions 0.101.0 through 0.103.11 and 0.104.0-alpha releases before 0.104.0-alpha.6 ignore X.509 name constraints that apply to URI names, causing such constraints to be accepted rather than enforced. Because name constraints are restrictions on otherwise properly issued certificates, the flaw is only reachable after successful signature verification and requires a misissued certificate to exploit; the library also provides no API for asserting URI names, and URI name constraints are otherwise unimplemented. Versions 0.103.12 and 0.104.0-alpha.6 reject URI name constraints unconditionally. |
| ArcadeDB before 26.9.1 contains an access control bypass vulnerability in the query_database tool exposed through the AI chat endpoints. The tool executes queries without binding the authenticated principal to DatabaseContext, causing per-type and per-bucket ACL checks to silently no-op and allowing authenticated users to read data they are explicitly denied at the per-type level. Attackers can prompt the AI assistant to execute queries against restricted types or buckets to retrieve sensitive data that would be rejected through normal query endpoints. |
| ImageMagick before 7.1.2-31 contains a policy bypass vulnerability in the UHDR encoder that fails to perform policy checks during buffer allocation for image pixels. Attackers can bypass resource policies by processing specially crafted UHDR images, potentially causing denial of service through excessive memory allocation. |
| A flaw was found in Netty's HTTP/2 stack. This vulnerability allows a remote attacker to inject prohibited characters, such as NUL, Line Feed, and Carriage Return, into HTTP/2 header field values due to insufficient validation. When these values cross an HTTP/2 to HTTP/1.1 translation boundary, they can be exploited for request smuggling, header injection, or response splitting. This could lead to unauthorized access, data manipulation, or other security bypasses. |
| A flaw was found in Netty's HTTP/2 codec. When converting HTTP/1 CONNECT requests to HTTP/2, the component incorrectly uses the Host header instead of the CONNECT authority-form request-target for the tunnel authority. A remote attacker can exploit this by supplying a different Host header, leading to a malformed HTTP/2 CONNECT request. This can bypass security controls such as tunnel allow-lists or egress policies, resulting in integrity loss. |
| A flaw was found in Netty RtspDecoder. The `RtspMethods.valueOf()` function incorrectly strips trailing control bytes from method tokens in Real-Time Streaming Protocol (RTSP) requests. A remote attacker can exploit this by sending a specially crafted RTSP request, leading to method-token smuggling. This vulnerability allows an attacker to bypass method-based access controls and can also be used to launder malicious requests through Netty-based RTSP proxies, making them appear legitimate to backend systems. |
| django-page-cms through 2.0.13 exempts five admin mutation views from CSRF protection in pages/admin/views.py, allowing attackers to forge requests that modify page content. Signed-in editors visiting a malicious page can be tricked into storing unescaped content that renders to all visitors, enabling stored cross-site scripting attacks. |
| snappy-java through 1.1.10.8 contains a buffer overflow vulnerability in typed Snappy.uncompress*Array methods that allocate output arrays by dividing uncompressed length by element size but pass the undivided length to native code. Attackers controlling compressed input can cause misaligned length values to write past array bounds with attacker-controlled bytes, corrupting heap memory. |
| NetworkManager-l2tp contains an improper input validation vulnerability that allows local users with VPN connection creation permissions to inject arbitrary pppd directives by supplying mru or mtu property values containing trailing non-numeric content after a valid integer. Attackers can exploit the verbatim write of unvalidated strings into the pppd options file via write_config_option() to inject the plugin directive, causing the privileged pppd process to load an attacker-controlled shared object and achieve arbitrary code execution as root. |
| SigNoz versions from 0.88.0 before 0.142.1 contain a SQL injection vulnerability in trace-funnel analytics endpoints that interpolate service_name and span_name fields into ClickHouse string literals without escaping. Authenticated attackers can inject SQL through funnel step definitions to execute arbitrary queries and read results in HTTP responses. |
| RAGFlow through 0.27.2 contains a path traversal vulnerability in the dev_insert_chunks_from_file and dev_insert_metadata_from_file endpoints that allows authenticated attackers to read arbitrary files by supplying absolute file paths in the file_path parameter. Attackers with valid access tokens can exploit missing path validation to read any file accessible to the service, with disclosure limited to files matching expected JSON structures that are then written to datasets. |
| Email::Sender::Transport::Sendmail versions before 2.602 for Perl allow arbitrary command execution on Windows sending a message whose envelope address reaches the shell in _sendmail_pipe.
On MSWin32 the envelope sender and every recipient go into a single command string, which open() passes to a shell. Every other platform gets the list form, which runs sendmail directly. When the caller supplies no envelope, Email::Sender::Simple takes the recipients from the To and Cc headers and the sender from the From header.
An attacker who controls one of those header addresses runs commands as the sending process. |
| HUBzero CMS through 2.2.32 accepts session identifiers from query strings and request variables instead of cookies alone, allowing unauthenticated attackers to fixate victim sessions. Attackers can obtain a valid session identifier, send victims a crafted link containing it, and replay the identifier after the victim authenticates to hijack their account and access. |
