vLLM versions 0.22.0 through 0.23.0 fail to validate stop_token_ids against vocabulary bounds in Rust HTTP and gRPC frontends, allowing out-of-vocabulary token IDs to reach MinTokensLogitsProcessor. Attackers can submit requests with min_tokens greater than zero and out-of-vocabulary stop_token_ids to trigger CUDA tensor indexing failures that leave EngineCore in a fatal state requiring service restart.
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.
vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths — the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route — the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact.
vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits and exhaust unaccounted GPU memory.
vllm before 0.29.0 fails to enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB limit in multimodal chat audio decoding, allowing unauthenticated clients to bypass file size restrictions. Attackers can submit oversized audio files through chat endpoints to consume excessive memory and CPU resources during decoding.
vLLM versions before 0.29.0 contain a denial-of-service vulnerability in the cache_salt parameter accepted on OpenAI-compatible and Anthropic API endpoints, which lacks maximum length validation and is processed on the single EngineCore scheduler thread. Unauthenticated attackers can send HTTP requests with multi-hundred-megabyte salt values that trigger expensive pickle serialization and SHA-256 hashing, stalling the scheduler thread and denying service to all concurrent requests.
SiYuan note-taking application versions 3.7.0 up to but not including 3.8.4 contain a stored cross-site scripting (XSS) vulnerability. This flaw occurs in the gallery and kanban database renderers where field descriptions are not properly escaped in aria-label attributes. In the Electron desktop app with nodeIntegration enabled, this allows attackers to inject JavaScript that can execute arbitrary commands with user privileges via Node.js child_process APIs. The vulnerability has a high severity score of 8.6.
SiYuan note-taking software versions before 3.8.4 contain a high-severity SQL injection vulnerability in the graph query endpoint. The vulnerability arises because the dailyNoteSavePath parameter is concatenated into an SQL query without proper escaping. This allows unauthenticated attackers on published sites with authentication disabled to perform SQL injection via UNION SELECT, potentially extracting arbitrary database rows from all notebooks.
SiYuan versions before v3.8.4 fail to properly escape four stored Attribute View values in textarea elements, allowing authenticated attackers to inject JavaScript by modifying field descriptions, template sources, select option descriptions, or footer calculation templates. Attackers can execute stored JavaScript when other users open affected database menus, and in the Electron desktop app with nodeIntegration enabled, this leads to command execution with SiYuan process privileges.
SiYuan versions from v2.1.0 before v3.8.4 contain a cross-site request forgery vulnerability in the CheckAuth lock-screen pass-through branch that grants administrator access to loopback requests without validating Origin headers. Attackers can craft malicious web pages that force victims to terminate the kernel process, read workspace configuration and proxy settings, and trigger administrative actions via zero-credential cross-origin requests from the victim's browser.
SiYuan note-taking software versions before 3.8.4 contain a cross-site scripting (XSS) vulnerability due to improper HTML escaping of flashcard block content. This flaw allows attacker-supplied content to execute scripts in a privileged renderer context when an administrator views the card manager. The vulnerability requires authentication and administrator privileges to exploit and can lead to arbitrary code execution on the host system.
SiYuan versions before 3.8.4 have an authorization omission vulnerability in the siyuan-get IPC handler. This flaw allows remote renderer processes to access native clipboard data formats such as MathML, Office, and WPS content during user-mediated paste operations. Attackers controlling remote renderer content can exploit this to read sensitive clipboard data without proper authorization.
SiYuan v3.8.3 fails to HTML-escape the data-subtype attribute when generating gutter-button markup (app/src/protyle/gutter/button.ts, assigned via innerHTML in app/src/protyle/gutter/index.ts) from content pasted as plain-text Markdown containing a Kramdown inline attribute list (IAL). Because the shared Lute renderer parses Kramdown IAL from text/plain input, an attacker-supplied Markdown snippet using entity-encoded quotes in data-subtype breaks out of the attribute value when the gutter markup is re-parsed by the browser, injecting additional attributes such as autofocus and onfocus. If a victim pastes the crafted Markdown and the affected gutter control receives focus, the injected handler executes; in the Electron desktop application, where the main BrowserWindow enables Node integration and disables context isolation, this results in JavaScript execution with renderer Node.js privileges (remote code execution). Fixed in v3.8.4.
SiYuan versions before v3.8.4 contain a path traversal vulnerability in the setNotebookIcon endpoint that allows authenticated administrators to create arbitrary directory trees and write files outside the workspace boundary. Attackers can supply directory traversal sequences in the notebook parameter to escape the workspace data directory and write conf.json files to arbitrary locations accessible by the kernel process.
SiYuan versions before 3.8.4 have a path traversal vulnerability in the checkoutRepo endpoint. Authenticated administrators can exploit this flaw by supplying a crafted sessionID parameter to write JSON files outside the intended workspace directory. This allows overwriting arbitrary JSON files in kernel-writable directories beyond workspace boundaries.
SiYuan versions before v3.8.4 contain a path traversal vulnerability in the exportBrowserHTML endpoint that allows authenticated administrators to write arbitrary HTML content to index.html outside the workspace directory. Attackers can supply a folder parameter with directory traversal sequences to escape the export directory and overwrite index.html in any pre-existing kernel-writable location, enabling stored XSS or workspace defacement.
SiYuan versions before 3.8.4 have an authentication bypass vulnerability in the publish service. This occurs because session cookies are issued without Secure or SameSite attributes over plaintext HTTP, allowing an attacker on the network path to capture and replay a valid session cookie. This enables unauthorized access to authenticated publish endpoints without needing the account password.
SiYuan note-taking software versions before 3.8.4 contain a missing authorization vulnerability in the 'siyuan-send-windows' IPC handler within the Electron main process. This flaw allows a renderer connected to a remote kernel to send commands that cause windows from other workspaces to lock repeatedly, resulting in a limited denial of service. There is no observed impact on confidentiality, integrity, or code execution.
SiYuan versions 3.8.0 through 3.8.3 contain an authorization bypass vulnerability in the MCP file tool's sensitive-path guard. The guard is applied only to the root of recursive operations but not to each descendant path, allowing an authenticated administrator to bypass protections on sensitive files and directories. This enables actions such as reading, copying, and overwriting protected files without proper confirmation. The issue is fixed in version 3.8.4.
Parse Server is an open-source backend server. In versions >= 9.0.0 and < 9.10.1-alpha.8, and in versions < 8.6.89, LiveQuery evaluates the protectedFields class-level permission against an incompletely resolved caller identity: the subscriber's roles are not resolved, and when a subscription does not supply its own session token the event payload is redacted against an anonymous identity even though the read was authorized against the connected user. As a result, field masks defined for a role, for authenticated users, or for a specific user are not applied, so an authenticated subscriber can receive field values that the REST API correctly withholds and can use a masked field to filter or watch a subscription. Only classes with LiveQuery enabled that define protectedFields under a role:, authenticated, or per-user group are affected; masks under the public (*) group are applied correctly. The issue is fixed in 9.10.1-alpha.8 and 8.6.89. As a workaround, additionally define the affected field masks under the public (*) group, or disable LiveQuery for classes whose class-level permissions rely on role-scoped, authenticated, or per-user protectedFields groups.