CVE Vulnerability Catalog

Apache NiFi 2.9.0 through 2.11.0 provide Connector configuration update and verification REST API methods that do not enforce authorization checking on Assets and Secrets referenced in the proposed configuration. An authenticated user authorized to modify a Connector but not to read a referenced Parameter Provider could apply Secret values backed by that Parameter Provider, and the methods accepted Asset identifiers without verifying ownership by the Connector being configured.

AsyncSSH is a Python package which provides an asynchronous client and server implementation of the SSHv2 protocol on top of the Python asyncio framework. Prior to 2.24.0, _process_channel_open and _process_channel_open_confirmation in asyncssh/connection.py accept a peer-supplied send_pktsize value of zero. When channel data reaches SSHChannel._flush_send_buf in asyncssh/channel.py, the zero value causes each loop iteration to slice and remove zero bytes without reducing the send window, leaving the synchronous loop permanently true with no await point. A malicious SSH server can trigger the client path through SSH_MSG_CHANNEL_OPEN_CONFIRMATION before the first channel write, while an authenticated client can trigger the server path through SSH_MSG_CHANNEL_OPEN and freeze every current and future connection handled by the process. This vulnerability is fixed in 2.24.0.

A vulnerability was found in Open5GS up to 2.8.0 affecting the function ogs_pfcp_parse_volume_measurement in lib/pfcp/types.c of the PFCP Handler component. The manipulation results in a null pointer dereference. The attack may be launched remotely, and the patch is identified as 8f07b507b78ff94776f2cd49276eb116ed93d7f2.

Svelte devalue is a JavaScript library that serializes values into strings when JSON.stringify isn't sufficient. Prior to 5.9.2, devalue.parse does not reject out-of-bounds indices greater than or equal to values.length in src/parse.js. A specially crafted untrusted payload can make the parser alternate between array representations, producing quadratic work and causing denial of service in applications that parse untrusted devalue data.

oRPC is a tool that helps build APIs that are end-to-end type-safe and adhere to OpenAPI standards. Prior to 1.14.8, the @orpc/server CORS plugin in packages/server/src/plugins/cors.ts copies a client's incoming Vary request header into the response instead of controlling Vary as a response-only header and using Origin for request-origin variation. In deployments behind a shared cache, CDN, or reverse proxy that keys responses using Vary, a client can inject arbitrary variation values, pollute cache keys, and cause inconsistent CORS enforcement for other clients.

Mattermost Desktop App versions <=6.2 6.2.2.0 did not sufficiently restrict server-rendered content from accessing local or private network resources. The issue was reported under the Mattermost responsible disclosure policy (MMSA-2026-00698).

On affected platforms running Arista EOS with IGMP snooping configured (enabled by default on all VLANs), a network-adjacent unauthenticated attacker can send malformed network packets on an affected VLAN to cause the IGMP snooping agent to terminate unexpectedly. This results in a temporary disruption of multicast traffic management, which may cause multicast traffic to be flooded to all ports of the affected VLAN until the service recovers.

Under certain circumstances on affected platforms running Arista EOS with gRPC Network Packet Sampling Interface (gNPSI) enabled, the gNPSI client credentials might be logged in clear text in local or remote accounting logs to authenticated users.

On affected platforms running Arista EOS with VRRPv2 IP-AH authentication configured, an unauthenticated attacker within the same layer 2 network segment on which VRRP is running can capture a legitimate authenticated VRRP advertisement and replay it indefinitely. Replayed advertisements can be used to advertise stale VRRP state, for example to prevent a backup router from taking over the virtual gateway after the original master has gone down.

Prior to @redocly/cli 2.33.2, the split command constructs output paths under --outDir from untrusted OpenAPI or AsyncAPI component names and x-codeSamples lang values without verifying that the resolved target remains inside the selected directory. A description containing literal ../ traversal segments in those fields can cause the split command to place or overwrite files outside --outDir.

Prior to 1.83.9, an authenticated LiteLLM Proxy caller with a valid virtual key can place api_base inside the user_config request body to bypass is_request_body_safe, which blocks top-level api_base and base_url but previously did not inspect or reject user_config. Because user_config constructs the outbound router, the nested destination redirects a server-side request to an internal or external host selected by the caller.

IRIS through 2.4.29 fails to properly validate case authorization in comment listing endpoints for notes, tasks, IOCs, assets, and evidence items. Attackers with access to any single case can enumerate sequential object identifiers and read comment threads from cases they have no authorization to access.

A vulnerability has been found in Open5GS up to 2.8.0 in the function smf_n4_handle_session_report_request of the file src/smf/n4-handler.c of the PFCP Session Report Request Handler component. The manipulation leads to a reachable assertion, and the attack may be initiated remotely.

A flaw has been found in Artifex MuPDF up to b6d17493700c621c0e70036980a6ebd06d2202c9 in the function pdf_open_filter of the file pdf-stream.c of the PDF Xref Loading component. Executing a manipulation can lead to a null pointer dereference, and the attack can be launched remotely.

kkFileView 5.0.0 through 5.0.2 allows reflected XSS via the /onlinePreview endpoint. The OnlinePreviewController passes the user-controlled page and kkagent request parameters to FreeMarker templates without sanitization, and the templates insert these values into raw JavaScript contexts.

Adobe Experience Manager is affected by a stored Cross-Site Scripting (XSS) vulnerability that could be abused by a low-privileged attacker to inject malicious scripts into vulnerable form fields. Malicious JavaScript may be executed in a victim's browser when they browse to the page containing the vulnerable field.

Prior to 0.28.0, request bodies for Chat Completions and Responses can set media_io_kwargs.video.video_backend to pynvvideocodec, and MediaConnector.fetch_video forwards that choice to VideoMediaIO even when startup configuration selected a software decoder. The engine's _reserve_mm_ipc_gpu_memory logic budgets decoder memory only from static configuration, so the request-selected VIDEO_LOADER_REGISTRY backend can create a CUDA context, decoder surfaces, and decoded-frame allocations that were not removed from the engine's KV-cache budget.

Concrete CMS before 9.5.3 exposed a legacy Express entry search endpoint that returned entry result JSON without invoking the canViewExpressEntries() permission check applied by the normal dashboard and CSV Export flow. An unauthenticated visitor who knew or discovered an Express entity identifier could enumerate that entity's entry search results, disclosing attribute values intended to be restricted to privileged users.

ContiNew Admin through 4.1.0 contains an authorization bypass vulnerability in the personal message delete endpoint that allows authenticated users to delete other users' messages and announcements. Attackers can supply arbitrary message identifiers in the IdsReq parameter to remove any message row and purge all recipients' read receipts without ownership validation.

Guns through 8.3.5 contains an improper access control vulnerability in SysNoticeController where requiredPermission defaults to false and is not overridden by any action methods. Authenticated users without assigned roles can exploit this to create, edit, delete, publish and retract system-wide notices affecting arbitrary users and departments.