CVE-2026-84286 is a critical out-of-bounds read vulnerability in Turboderp-org's ExLlamaV3 version 3. The EXL3 loader uses an untrusted model-supplied index to select a CUDA kernel without verifying bounds, allowing a malicious model to cause a crash or potentially achieve controlled indirect kernel dispatch under specific memory conditions.
In Progress ShareFile Storage Zones Controller v5.12.6 and below, improper validation of outbound request destinations and parameters, unsafe JSONP response formatting, and insufficient validation on queue service can be chained by a remote unauthenticated attacker to issue server-side requests to unintended local handlers and perform unauthorized queue operations.
CVE-2026-21132 is an improper input validation vulnerability in Samsung Internet browser versions prior to 30.0.5.24. This flaw allows remote attackers to inject arbitrary scripts, but requires user interaction to trigger. The vulnerability has a medium severity rating with a CVSS 4.0 score of 6. Affected versions are those before 30.0.5.24. No explicit patch information is provided in the input data.
CVE-2026-21130 is a medium severity vulnerability in Samsung Checkout prior to version 5.1.03.1. It involves improper input validation in the CreditCardRetrieveActivity component, which allows local attackers to launch arbitrary activities with Samsung Checkout privileges.
CVE-2026-21129 is a medium severity vulnerability in Samsung Checkout prior to version 5.1.03.1. It involves improper input validation in the GuestCreditCardRegisterActivity component, which allows local attackers to launch arbitrary activities with Samsung Checkout privileges.
CVE-2026-21128 is an improper input validation vulnerability in the CreditCardRegisterActivity component of Samsung Checkout prior to version 5.1.03.1. This flaw allows local attackers to launch arbitrary activities with Samsung Checkout privileges. The vulnerability has a medium severity score of 5.1 (CVSS 4.0). No known exploits or patches are currently documented.
CVE-2026-21126 is an improper input validation vulnerability in Samsung Mobile Devices affecting the libsimba.heifdec.media.samsung.so component prior to the SMR Oct-2026 Release 1. This flaw allows local attackers to write out-of-bounds memory, potentially leading to memory corruption. The vulnerability has a CVSS 4.0 base score of 6.9, indicating medium severity. No official patch or remediation details are provided in the available data.
CVE-2026-21125 is a high-severity vulnerability in Samsung Mobile Devices involving improper input validation in the parsing of 'sheifd' within the library libsimba.heifdec.media.samsung.so. This flaw affects versions prior to the Samsung Mobile Security Release (SMR) October 2026 Release 1. It allows local attackers to potentially execute arbitrary code on the device without requiring user interaction or elevated privileges.
CVE-2026-21124 is a high-severity vulnerability in Samsung Mobile Devices involving improper input validation in the component libsimba.heifdec.media.samsung.so. This flaw affects the process of obtaining cover images and can allow local attackers to potentially execute arbitrary code. The vulnerability exists in versions prior to the SMR Oct-2026 Release 1. No known exploits are reported in the wild, and no patch links are provided in the available data.
A vulnerability in Unisoc (Shanghai) Technologies Co., Ltd. modem products T8100, T9100, and T8300 allows local information disclosure due to missing filtering of sensitive data in crash dumps. This issue does not require additional execution privileges and has a medium severity rating with a CVSS score of 5.5.
CVE-2026-87765 is a medium severity vulnerability in the Unisoc kernel apsys DVFS driver caused by improper input validation of the frequency index. This flaw can lead to out-of-bounds reads or hardware register corruption, potentially causing system instability or denial of service when exploited with local privileges.
CVE-2026-107825 is an improper input validation vulnerability in Coraza WAF versions 3.0.0 through 3.7.x. The ProcessURI function fails to properly handle malformed URIs containing control bytes, causing query parameters to be omitted from internal variables while downstream integrations may still process them. This discrepancy allows unauthenticated attackers to bypass security rules targeting query parameters. The issue is fixed in version 3.8.0.
Apache Camel Karavan versions from 4.0.0 up to but not including 4.22.1 contain an improper input validation vulnerability. This flaw allows an authenticated user of any role to deploy arbitrary Kubernetes resources by submitting crafted kubernetes.yaml files without restrictions on resource kinds or security-sensitive pod options. The vulnerability enables the creation of pods with elevated privileges or access to sensitive host resources. The issue is fixed in version 4.22.1.
SumatraPDF is a multi-format reader for Windows. In 3.5.2 and earlier, an attacker-controlled SyncTeX source filename is substituted for the %f placeholder in an external editor command line without safe Windows argument quoting, and the resulting command line is passed to CreateProcessW(). A user with an external editor configured or auto-detected who opens a PDF with a crafted .synctex.gz file and invokes inverse search can inject command-line flags; the resulting impact depends on the target editor interpreting those flags and can include unintended editor actions or code execution through a malicious extension. No broader impact is claimed beyond the advisory-supported conditions. No fixed version is available as of this review.
fast-jwt provides fast JSON Web Token (JWT) implementation. Prior to 6.3.1, fast-jwt createVerifier accepts an unsigned JWT when key is an empty string or null and algorithms is a non-empty allowlist. Falsy synchronous keys bypass prepareKeyOrSecret, allowedAlgorithms remains active, hasKey is false, and the empty signature avoids the verifySignature gate. An attacker can therefore submit a token containing arbitrary claims without possessing a signing key, resulting in authentication or authorization bypass. Claim validators still run, and non-empty keys, an empty key without algorithms, and the async key resolver path do not have this behavior. This issue is fixed in version 6.3.1.
CVE-2026-107717 is an improper input validation vulnerability in masci banks prior to version 2.5.0. The Banks Prompt.chat_messages() function parses each line of rendered template output as ChatMessage JSON. If untrusted data is rendered and passed to an LLM provider, attacker-controlled JSON can manipulate the ChatMessage.role field, allowing an attacker to override application instructions or alter prompt structure. This issue is fixed in version 2.5.0.
Improper Input Validation vulnerability in BeamMCP.Schema in ScriptKittyOS beam_mcp allows an MCP client to reach a tool's dispatch function with arguments that violate the input schema the server advertised. BeamMCP.Schema.validate/2 checked type, required, additionalProperties, enum and numeric bounds on the top-level arguments object only. Constraints inside nested objects and on array items (items, minItems, maxItems, minLength, maxLength, pattern, nested required, enum and additionalProperties: false) were advertised by tools/list and never checked at tools/call or prompts/get, and keywords outside the enforced subset (oneOf, anyOf, $ref) were advertised and ignored.
A host whose dispatch code relies on the schema it declared receives values the schema forbids, such as an out-of-range number or an undeclared key inside a nested object. What the host does with such a value decides the impact.
This issue affects beam_mcp: from 0.1.0 before 0.10.1.
The Wallet System for WooCommerce WordPress plugin before 2.8.0 does not validate that a wallet transfer amount is positive, and computes the sender's new balance from a stale snapshot taken before crediting the recipient, allowing an authenticated attacker with Subscriber-level access to mint wallet funds for themselves or drain a specific victim's balance into their own account.
CVE-2026-107225 is an improper input validation vulnerability in the Go library excelize versions 2.8.0 through 2.11.0. The vulnerability arises because the GetStyle function's predicates only check upper bounds for style-table indices, allowing negative indices to be used. When a crafted styles.xml file contains negative fillId, borderId, or fontId values, these negative values bypass the upper-bound checks and cause negative slice indexing, leading to a panic during cell styling reads. No fix is currently available for this issue.
CVE-2026-106443 is a high-severity remote code execution vulnerability in Kozea WeasyPrint that arises from improper input validation. WeasyPrint passes untrusted image bytes directly to Pillow's format dispatcher without restricting input formats. When Ghostscript is installed on the host, Pillow's EPS image plugin invokes Ghostscript to rasterize EPS/PS input, allowing attacker-controlled PostScript to execute. Hosts running vulnerable Ghostscript versions with known -dSAFER bypasses can be fully compromised. The vulnerability affects WeasyPrint versions prior to 70.0 and impacts any deployment rendering untrusted content on hosts with Ghostscript installed.