A vulnerability was detected in crossplane crossplane-runtime up to 2.2.2/2.3.2. This vulnerability affects the function Get of the file pkg/xpkg/client.go of the component ImageConfig. The manipulation results in time-of-check time-of-use. The attack may be launched remotely. Upgrading to version 2.2.3, 2.3.3 and 2.4.0-rc.1 is able to resolve this issue. The patch is identified as bee99c6cd6ca81878acca2940a2f0a02169fc208. You should upgrade the affected component.
CVE-2026-103602 is a vulnerability in the Legion of the Bouncy Castle Inc. bc-csharp library before version 2.7.0. It involves improper certificate validation in the PkixNameConstraintValidator component, where trailing dots in fully qualified host names were not properly handled. This flaw allows an attacker controlling or able to obtain certificates from a name-constrained intermediate CA to bypass name constraints during PKIX certification path validation for email addresses, DNS names, or URI hosts within excluded subtrees.
In Bouncy Castle for Java before 1.86, NTRU reduced secret values with the % operator in three helpers whose reference implementations are deliberately division-free, so each reduction was carried out by an integer division whose latency depends on the secret operand. Polynomial.modQ divided by a variable divisor, which a compiler cannot strength-reduce to a multiply the way it can a constant one, so it emitted a division on every call including on the decapsulation path where the dividend derives from the private key; Polynomial.mod3 and NTRUSampling.mod3 divided the secret key polynomials f and g during key generation, the message polynomials r and m during encapsulation, and coefficients recovered during decapsulation. An attacker able to measure that timing can recover information about the NTRU private key. modQ now masks, which is exact because q is always a power of two, and mod3 uses the reference implementation's division-free fold and select; the results are unchanged.
Discord libdave before 1.2.0 did not reject an MLS Welcome message when the resulting group roster contained an unrecognized participant. An attacker in control of the DAVE signaling path (the voice gateway, or an equivalent position able to add, alter, or withhold signaling messages to a client) could cause affected clients to accept an unauthorized member into the end-to-end encrypted media session, compromising the confidentiality and integrity of audio and video.
Improper Verification of Source of a Communication Channel in the ADS discovery of the Go implementation of Apache PLC4X (PLC4Go) allows an attacker able to send UDP datagrams to the discovering host to redirect subsequent connections to an arbitrary, attacker-chosen address. The discovery result's connection
address was derived from the AmsNetId claimed in the response body rather than from the datagram's actual source address. One spoofed discovery response can therefore insert an inventory entry pointing at any host, including hosts outside the local network, and an application that connects to discovered devices
will open its ADS session, including any configured route credentials, to that host.
Additionally, discovery listeners in both implementations can be disabled by a single malformed datagram:
- In PLC4Go ADS discovery, a short version block causes a panic that ends the listener for the rest of the discovery call, so legitimate devices answering afterwards are not reported.
- In PLC4J, the ADS and EtherNet/IP discoverers stop on an unhandled exception from a malformed response.
- The PLC4J Modbus discoverer can be made to spin indefinitely, consuming a CPU core, by a scanned host that sends a partial response.
Exploitation requires the application to invoke the discovery API, which is opt-in, and for the connection redirect, to act on the discovered items.
This issue affects Apache PLC4X: PLC4Go from 0.11.0 before 1.0.0; PLC4J ADS and Modbus drivers from 0.10.0 before 1.0.0; PLC4J EtherNet/IP driver from 0.11.0 before 1.0.0. PLC4Go is consumed as the Go module github.com/apache/plc4x/plc4go; versions refer to the corresponding Apache PLC4X releases.
Users are recommended to upgrade to version 1.0.0, which fixes the issue. Version 1.0.0 derives the connection address from the datagram's source address and logs a warning when the claimed AmsNetId disagrees with it.
Integer Overflow, Improper Validation of Array Index, Uncontrolled Recursion and Memory Allocation with Excessive Size Value in the Go implementation of Apache PLC4X (PLC4Go) allow a malicious device, or an attacker able to inject network traffic, to crash or exhaust the memory of the client application,
causing a denial of service.
The individual defects are:
- Generated parsers pre-allocate arrays with the element count claimed on the wire (0.13.0 through 0.13.1).
- Transport read helpers allocate buffers of the size claimed on the wire without an upper bound.
- ADS and KNXnet/IP response handling indexes into received data without checking its length, causing a panic.
- ADS and EIP frame-length handling accepts, or arithmetically wraps to, a length of zero, breaking message framing.
- Recursive protocol types are parsed without a nesting-depth limit. The same defect in the Java implementation is covered by CVE-2026-102509 https://cveprocess.apache.org/cve5/CVE-2026-102509 .
Additionally, length and position arithmetic in generated serializers was performed in 16-bit integers. If an application forwards attacker-influenced payloads larger than 8 KB, the length field wraps, and the remainder of the payload may be interpreted by the receiving device (for example, an ADS PLC) as
additional, independent protocol messages.
This issue affects Apache PLC4X: from 0.11.0 before 1.0.0. PLC4Go is consumed as the Go module github.com/apache/plc4x/plc4go; versions refer to the corresponding Apache PLC4X releases.
Users are recommended to upgrade to version 1.0.0, which fixes the issue.
A flaw has been found in zosmaai pi-llm-wiki up to 0.11.7. Affected is an unknown function of the file mcp/index.ts of the component wiki_capture_source MCP tool. Executing a manipulation of the argument url can lead to os command injection. The attack can be executed remotely. The exploit has been published and may be used. Upgrading to version 0.11.8 is able to address this issue. This patch is called 360867034e79175b45c8e04a98e4ca712bbaca35. Upgrading the affected component is advised.
The device's update mechanism retrieves metadata for software updates over an unencrypted HTTP connection and stores portions of that metadata for later use. A management interface subsequently returns this stored value in a JSON response, and the web interface responsible for displaying update information inserts that value directly into the page as HTML. This behavior allows attacker‑controlled metadata to be interpreted as script content. In addition, the same authenticated origin provides an interface capable of executing system‑level commands with root privileges. An attacker able to influence update metadata could exploit these conditions to execute arbitrary code within the administrative context of the device.
IBM Guardium Data Protection version 12.2 contains a command injection vulnerability in its SNMP alert notification feature. An authenticated attacker able to modify policy alert text can execute arbitrary operating system commands with root privileges via the SNMP alerter service. This vulnerability is identified as CWE-78 and has a CVSS 3.1 score of 7.5, indicating high severity.
Issue summary: The OpenSSL QUIC server, when configured to not preform address validation, can be forced to count incoming packets multiple times in its unvalidated credit computation, leading to a violation of the RFC 9000 unvalidated connection amplification limit of 3 times the amount of data received. Impact summary: A remote attacker able to spoof packets to a server using the OpenSSL QUIC implementation might use the server for an amplification of a DDoS attack. CWE: CWE-440: Expected Behavior Violation Description: OpenSSL's QUIC stack, when operating as a server, enforces client address validation (RFC 9000, Section 8), to confirm the peer address is not used for a traffic amplification attack. If this feature is disabled on the server, the QUIC stack limits the amount of server data that can be sent to 3 times the amount of data received from the peer address, until such time as the TLS handshake is completed. The OpenSSL QUIC server, when operating in non-validation mode, adds the length of the whole datagram received to the unvalidated credit limit when processing each QUIC packet in the datagram. A remote peer may, after establishing a connection with an initial client hello frame, send a subsequent datagram containing multiple QUIC packets, leading the server to account the entire datagram length for each packet in the datagram, resulting in the server believing that the peer has sent more data than it actually has, thereby violating the 3x amplification limit mandated by the RFC. FIPS impact: no As the QUIC stack lives outside the FIPS module boundary, no FIPS modules are affected by this CVE.
FreePBX is an open source IP PBX. Prior to version 17.0.9, authenticated users who are authorized to access the GraphQL api module interface of FreePBX are able to execute arbitrary shell commands. Authenticated access to the api module is required. The PBX API module's documentation generator accepts an authenticated host parameter and uses it to build a shell command. The code path validates the generated OAuth access token before execution, but it does not validate or escape host. Compromise results in authenticated arbitrary shell command execution as the FreePBX web/PBX service user (typically asterisk.). This issue has been patched in version 17.0.9.
A flaw has been found in ag-ui-protocol ag-ui up to 2026-09-07. This vulnerability affects the function FilterToolCallsMiddleware of the file sdks/typescript/packages/client/src/middleware/filter-tool-calls.ts of the component Middleware. Executing a manipulation can lead to incomplete cleanup. The attack may be launched remotely. Upgrading to version 2026-09-08 is able to resolve this issue. This patch is called c346119fe870b70f5c19738ee5119f3e1456e59d. It is suggested to upgrade the affected component.
A vulnerability was identified in Tencent AI-Infra-Guard up to 4.5.2/4.6.2. This affects the function startsWith of the file skill_scan/tools/dir/dir_actions.py of the component File Access. The manipulation leads to path traversal. The attack needs to be performed locally. The exploit is publicly available and might be used. Upgrading to version 4.6.0 is able to mitigate this issue. The identifier of the patch is ac0384edc9dbea3b226edefcf50613bd8509134f. You should upgrade the affected component.
A vulnerability was detected in October CMS up to 4.3.4. This affects the function validateExternalImageHost of the file System/Classes/ResizeImages.php of the component SSRF Protection. The manipulation results in server-side request forgery. The attack may be launched remotely. The exploit is now public and may be used. Upgrading to version 4.3.5 is able to mitigate this issue. You should upgrade the affected component.
Contrast is a confidential-computing runtime for Kubernetes. In versions before 1.19.1, the Kata agent policies generated by the Contrast CLI contained a flaw in the CopyFile verification that allowed arbitrary writes to the guest root filesystem. A malicious process on the untrusted host able to connect to the Kata agent VSOCK could issue a series of CopyFile requests to overwrite security-critical files in the guest or trick the workload into disclosing sensitive data, effectively amounting to a full guest takeover. Users unable to upgrade can apply an equivalent rego policy fix passed to 'contrast generate --policy'.
Contrast, Edgeless Systems' runtime for confidential containers on Kubernetes, is affected in versions up to and including 1.9.0. The VOLUME directive in a Dockerfile (config.volumes in the OCI image configuration) is only a hint and is not handled specially by Kubernetes, but containerd adds a mount point for it when Kubernetes sets none, requiring the runtime to be able to push arbitrary data to the Kata agent. As a result, on bare-metal Contrast deployments (AKS deployments are not affected) that run an image declaring at least one VOLUME for which no Kubernetes mount exists at that path, the untrusted host can write arbitrary file trees below that mount point inside the confidential container, compromising the integrity of a directory that is typically important to the application's core functionality. Version 1.9.1 fixes the issue by disallowing this configuration in `contrast generate`.
Froxlor versions up to and including 2.3.10 have a critical vulnerability where the system.letsencryptchallengepath setting is not properly restricted or escaped. This allows an administrator or any actor able to write settings to inject arbitrary command-line options into the acme.sh command executed by root via cron, leading to arbitrary command execution or file writes as root. The issue is fixed in version 2.3.12.
vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate. When the request contains a 'features' (multimodal) payload, vllm/entrypoints/serve/disagg/serving.py builds a multimodal EngineInput directly from the caller-supplied token_ids, and GenerateRequest.token_ids (vllm/entrypoints/serve/disagg/protocol.py) is not checked against model_config.max_model_len. For multimodal processors that report skip_prompt_length_check=True (for example Nemotron Parse, Whisper, and FireRedLID), InputProcessor._validate_prompt_len() returns immediately for both encoder and decoder prompts, so an overlong prompt becomes an EngineCoreRequest and reaches the worker input-batch copy into a fixed max_model_len-wide NumPy row. A client able to reach the endpoint on an affected model configuration can therefore submit an overlong token_ids list to trigger a worker failure and denial of service. Fixed in 0.29.0.
OpenClaw versions 2026.5.1 through 2026.7.0 fail to apply the configured exec approval path to Google Meet node commands. The googlemeet.chrome command accepts caller-supplied audio command arrays and executes them on a paired node without going through the normal system.run approval flow. In deployments with the Google Meet plugin enabled, a paired Chrome node, and the googlemeet.chrome node command allowed, a tool-enabled agent able to invoke that command can execute attacker-selected processes on the paired node, impacting files, credentials, browser profiles, and availability on that node. The issue is fixed in 2026.7.1; as a workaround, remove googlemeet.chrome from allowed node commands or disable the Google Meet plugin.
OpenClaw (npm package 'openclaw') before 2026.7.1 is vulnerable to a time-of-check time-of-use race condition in OpenShell local mirror filesystem mutation operations. The remove, mkdir, and rename operations could act on a different filesystem target after OpenClaw completed its sandbox path-safety check, if the path is changed concurrently. An attacker able to win the race can cause a sandboxed operation to delete, create, or rename a host path outside the intended mirror root with the permissions of the OpenClaw process user. This does not require an operator to have granted host filesystem access outside the sandbox. The issue is fixed in 2026.7.1.