CVE Watch — Your Vulnerability Feed

Rclone before 1.75.2 contains a path traversal vulnerability in the local backend's translateLink function, which trims the .rclonelink suffix from a path after the containment check, so a name can resolve outside the destination root. Attackers can plant a directory named ...rclonelink in a source tree synced with --links and --metadata to change ownership, timestamps, and xattrs of the destination's parent directory.

Snipe-IT before 8.6.3 ignores the ldap_start_tls() return value in Ldap::connectToLdap(), allowing on-path attackers to force cleartext LDAP binds. Attackers can answer the StartTLS request with a protocol error to capture service-account credentials and end-user login passwords.

rclone from 1.37 before 1.74.2 contains a server-side request forgery vulnerability that allows unauthenticated attackers to redirect http and netstorage backend requests by supplying a double-slash root on an existing named remote via --rc-serve. Attackers can make HEAD/GET requests to attacker-chosen hosts to read responses, capture configured custom headers, or obtain valid NetStorage request signatures.

vLLM from 0.20.0 before 0.30.0 contains a denial of service vulnerability that allows remote attackers to crash the engine by sending inconsistent multimodal metadata to the disaggregated /inference/v1/generate endpoint. Attackers can replay a valid multimodal item with repeated or overlapping mm_placeholders ranges or mismatched kwargs_data and mm_hashes lists, producing a negative text span in Qwen2.5-VL position construction that terminates the engine.

vLLM from 0.22.0 before 0.24.0 contains an improper input validation vulnerability in the Rust gRPC converter that allows remote attackers to bypass max_logprobs limits via CandidateTokens.top_n. Attackers can send Generate or GenerateStream requests with oversized or integer-wrapping top_n values to trigger torch.topk failures that terminate EngineCore.

vLLM from 0.22.0 before 0.24.0 contains a denial of service vulnerability that allows remote attackers to terminate EngineCore by bypassing logprob_token_ids limits in the Rust gRPC frontend. Attackers can send Generate or GenerateStream requests with output_logprobs enabled and more than 128 output_candidates.token_ids, triggering a fatal sampler exception that crashes the engine.

vLLM from 0.22.0 before 0.24.0 contains an improper input validation vulnerability in the Rust gRPC Generate and GenerateStream service that forwards unchecked prompt token IDs to EngineCore. Attackers can submit requests with out-of-vocabulary prompt.token_ids values to trigger a fatal embedding lookup failure that terminates EngineCore and leaves the server unhealthy.

ProjectSend 2.0.0 through 2.7.1 contains an incomplete authorization vulnerability in FileVersions::link() that allows staff users to share files without authorization by linking file versions. Attackers can link their file onto an intermediate revision or assign out-of-scope clients, causing moveAssignmentsToRoot() to expose protected original files to unauthorized clients.

rotki before 1.44.2 contains a weak key derivation vulnerability that allows attackers to brute-force backups because the premium sync key is a single unsalted SHA-256 of the user password. Attackers who obtain an uploaded backup blob can run GPU-parallel offline password guessing and decrypt the database containing exchange API keys, secrets, wallet addresses and transaction history.

A vulnerability was determined in NocoDB up to 2026.07.0. This vulnerability affects the function UserRefreshToken.deleteAllUserToken of the file /auth/user/signup of the component PublicApiLimiterGuard. This manipulation causes allocation of resources. The attack is possible to be carried out remotely. The attack is considered to have high complexity. It is stated that the exploitability is difficult. Code-verified absence of in-app rate limiting on the unauthenticated signup and password-forgot endpoints and per-request refresh/OAuth token invalidation, but the vendor documents the pass-through limiter as intentional design. The vendor deleted the GitHub issue for this vulnerability without any explanation. Afterwards the vendor was contacted early about this disclosure via email but did not respond in any way.

A vulnerability was found in NocoDB up to 2026.08.0. This affects an unknown part of the file packages/nocodb/src/services/users/users.service.ts of the component Password-Reset Flow. The manipulation results in race condition. The attack can be executed remotely. The vendor deleted the GitHub issue for this vulnerability without any explanation. Afterwards the vendor was contacted early about this disclosure via email but did not respond in any way.

Improper Neutralization of Special Elements used in an Expression Language Statement ('Expression Language Injection') vulnerability in Apache Camel AWS2-s3 component. This issue affects Apache Camel: from 4.8.0 before 4.18.5, from 4.19.0 before 4.22.2. For the object it uploads, reads or deletes, the camel-aws2-s3 producer takes the key from the CamelAwsS3Key message header and falls back to the keyName option configured on the endpoint only when that header is absent. Since 4.8.0, AWS2S3Utils.determineKey() also passed the resulting value to the Simple language whenever it contained ${ or $simple{, so that a route could use a dynamic key, and the evaluation applied to a value taken from the header as well as to the configured option. The header is not route configuration, however: the aws2-s3 consumer sets CamelAwsS3Key on the exchanges it creates to the key of the object it read, that is, the name of the object in the bucket. In a route that does nothing more than consume from one aws2-s3 endpoint and produce to another, the name of each consumed object therefore reached the producer, and a name containing either marker was evaluated as an expression, although the route author never asked for any evaluation, and configuring keyName on the producer did not prevent it, because the header took precedence. A key that a route copied into the header from other message content was evaluated in the same way. Anyone able to choose the name of an object in a bucket that such a route consumes, typically a principal allowed to upload to that bucket, could therefore have Simple expressions of their choosing evaluated in the Camel process. The Simple language can read JVM system properties, environment variables and configuration properties, which commonly carry credentials, and, where camel-bean is on the classpath, invoke methods on beans in the registry. The evaluated result became the key the producer used, so reading a value back additionally required the ability to

Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal') vulnerability in Apache Camel Docling component. This issue affects Apache Camel: from 4.15.0 before 4.18.5, from 4.19.0 before 4.22.2. DoclingProducer builds the command line of the external docling tool and appends to it the custom arguments supplied in the CamelDoclingCustomArguments header, a list of strings whose flags the fix for CVE-2026-40047 restricted to an allowlist. That fix also introduced validatePathSafety to keep path-like argument values inside the intended directory, but the method only looked for relative traversal: it rejected a value containing a traversal sequence such as ../ and a value whose normalized form still held a .. element. It never checked whether a value was an absolute path, and an absolute path without traversal segments contains neither, so it passed unchanged. Every path-like value went through this check, whether it was a separate list element following its flag or was attached to the flag after an equals sign, so an absolute path was accepted in either form, for example as the directory given to the allowlisted --artifacts-path flag, which tells docling where to find its model artifacts. Such a value reached the docling command line as given, so a party able to choose the custom arguments could point the docling process at any directory the Camel process can access, outside the working directory that the check was meant to keep path values in, which is the outcome the CVE-2026-40047 fix set out to prevent. Only the CLI mode, which is the default, passes custom arguments to docling. The header belongs to the Camel header namespace, which consumers applying Camel's default header filtering, such as the JMS and Kafka consumers, remove from incoming messages, so the issue is reached where a route forwards externally influenced values into the header or where a consumer maps incoming headers without that filtering. Users are recommended to upgra

Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal') vulnerability in Apache Camel Docling component. This issue affects Apache Camel: from 4.15.0 before 4.18.5, from 4.19.0 before 4.22.2. DoclingProducer runs the external docling tool in CLI mode, the default, and tells it where to write the converted document through the --output flag. When the CamelDoclingOutputFilePath header was set, addOutputDirectoryArguments handed its value to that flag verbatim: the value was not normalized, an absolute path or a path containing .. segments was accepted as given, and no endpoint option could confine it to a directory. The custom-argument validation added by the fix for CVE-2026-40047 refuses the --output and -o flags in the CamelDoclingCustomArguments header and tells callers to choose the output directory with this header instead, so the one value that decides where docling writes reached the command line without any of the checks applied to custom arguments. Docling writes the conversion result, as Markdown, HTML, JSON or text, into that directory, naming the file after the input file, and the producer afterwards removes only its own temporary directory, so the file is left in place. A party able to set the header could therefore have conversion output created in any directory the Camel process can write to. The content of such a file is the conversion of the document carried by the message, so it is shaped by whoever supplies that document without being written byte for byte, and when the document content arrives in the message body the producer stores it under a random temporary name, which the output file then carries. The header belongs to the Camel header namespace, which consumers applying Camel's default header filtering, such as the JMS and Kafka consumers, remove from incoming messages, so the issue is reached where a route forwards externally influenced values into the header or where a consumer maps incoming headers without that

A vulnerability was detected in Datafeedr Product Sets Plugin up to 1.3.24 on WordPress. Affected by this issue is the function dfrps_ajax_delete_saved_search of the file functions/ajax.php of the component AJAX Handlers. The manipulation results in missing authorization. The attack may be launched remotely. The exploit is now public and may be used. Upgrading to version 1.3.25 can resolve this issue. Upgrading the affected component is advised. The vendor was contacted early about this disclosure.

YesWiki before 4.7.0 contains a stored cross-site scripting vulnerability that allows low-privilege authenticated users to upload unsanitized SVG files through the bazar ImageField class. Attackers can submit a malicious SVG via the image field of the entries API, which is served as image/svg+xml and executes script for users viewing the entry or file.

SiYuan from 2.10.0 before 3.8.7 contains a stored cross-site scripting vulnerability that allows users with edit permission to inject script by placing javascript: URLs in math block content rendered with KaTeX trust enabled. Attackers can store \href, \url, or \includegraphics payloads that mathRender assigns via innerHTML, executing script in the application origin when a reader clicks the rendered link.

WWBN AVideo 26.0 through 29.3.0 contains a stored cross-site scripting vulnerability that allows authenticated commenters to inject JavaScript because markDownToHTML() preserves event-handler attributes in leading HTML blocks. Attackers can post a comment starting with an img tag carrying an onerror handler, which bypasses the inline sanitizer and executes in viewers' browsers under the application origin.

Infinity for Reddit through 8.3.4 contains a login cross-site request forgery vulnerability that allows attackers to log victims into an attacker-controlled account because the exported LoginChromeCustomTabActivity redirect handler checks a hardcoded OAuth state value. Attackers can authorize the app's client ID with their own Reddit account, then deliver an infinity://localhost redirect containing their code and the constant state, causing the victim's posts, messages, and browsing to occur on the attacker's account.

Missing Authorization vulnerability in VillaTheme Bopo – WooCommerce Product Bundle Builder bopo-woo-product-bundle-builder allows Exploiting Incorrectly Configured Access Control Security Levels.This issue affects Bopo – WooCommerce Product Bundle Builder: from n/a through 1.2.3.