Security Vulnerabilities
- CVEs Published In May 2026
Cargo between 1.68 and 1.96 incorrectly normalized the URLs of third-party registries using the sparse index protocol. If a hosting provider allowed multiple registries to be hosted with arbitrary names within the same domain, an attacker able to publish crates in a registry could obtain the credentials of others users of the same registry. The severity of the vulnerability is **low**, due to the extremely niche requirements needed to achieve the attack.
Cargo incorrectly handled symlinks inside of crate tarballs downloaded from third-party registries, allowing a malicious crate to override the source code of another crate from the same registry. The severity of the vulnerability is **medium** for users of third-party registries. Users of crates.io are **not affected**, as crates.io forbids uploading crates containing any symlink.
Apache Airflow providers-google's `ComputeEngineSSHHook` disables SSH host-key verification by default, exposing SSH traffic between an Airflow worker and a Compute Engine VM to in-path network attackers who can intercept or modify the session. Users are advised to upgrade to `apache-airflow-providers-google` 22.0.0 or later.
A security vulnerability has been identified in Acer Care Center where the ACCSvc service creates a Named Pipe with a weak Security Descriptor. This vulnerability allows an authenticated local user to connect and send a specially crafted message (message type 0x03) to the pipe, causing the service to crash with exit code 1067 (ERROR_PROCESS_ABORTED). To mitigate this potential local service disruption, Acer requires users to update the software to the latest version.
A cross-site scripting (XSS) vulnerability exists in Apache ECharts in the Lines series tooltip rendering logic.
This issue affects Apache ECharts: from before 6.1.0.
In versions prior to 6.1.0, if both Lines series and tooltip are used, and no user-specified tooltip.formatter is provided, and series.data[i].name is specified, raw HTML string series.data[i].name can be rendered through innerHTML sink into tooltip content. Although tooltip is allowed to accept user-provided raw HTML via a custom tooltip.formatter, the built-in tooltip formatters conventionally perform HTML escaping automatically. This case breaks that convention and may unexpectedly lead to script execution when tooltips are displayed.
Users are recommended to upgrade to version 6.1.0 if using the Lines series in this way, which fixes the issue.
Mattermost versions 11.6.x <= 11.6.0, 11.5.x <= 11.5.3, 11.4.x <= 11.4.4, 10.11.x <= 10.11.14 fail to filter nil elements from outgoing webhook attachment payloads before processing, which allows an authenticated user to cause a denial of service (server process termination) via a crafted webhook callback response containing a null attachment entry.. Mattermost Advisory ID: MMSA-2026-00641
Spring AI's support for Anthropic's Skills API used LLM-influenced filenames unsanitized in Path.resolve before writing files to disk. This could allow a malicious user to write files outside the intended target directory, including restricted directories.
Affected versions:
Spring AI: 1.1.0 through 1.1.x
A vulnerability in MLflow versions <=3.10.1.dev0 allows unauthorized access to multipart upload (MPU) endpoints when the `--serve-artifacts` mode is enabled. The authorization logic does not enforce resource-level permission checks for `/mlflow-artifacts/mpu/*` endpoints, enabling attackers to overwrite artifacts belonging to other users. This can lead to unauthorized cross-user writes, model supply chain poisoning, and arbitrary code execution when compromised models are loaded. The issue is resolved in version 3.10.0.
A critical remote code execution vulnerability exists in all versions of the HuggingFace transformers library prior to version 5.3.0. The vulnerability allows an attacker to craft a malicious `config.json` file containing the `_attn_implementation_internal` field set to an attacker-controlled HuggingFace Hub repository ID. When a victim loads this model using the standard `AutoModelForCausalLM.from_pretrained()` API, the library downloads and executes arbitrary Python code from the attacker's repository with the victim's full OS privileges. This issue arises due to unfiltered deserialization of configuration attributes, insufficient sanitization of internal fields, and unsandboxed execution of downloaded kernels. The vulnerability bypasses the `trust_remote_code` security mechanism, is invisible to the victim, and exploits the standard documented usage pattern, making it particularly severe. Users are advised to upgrade to version 5.3.0 or later to mitigate this issue.
Dolibarr ERP CRM 7.0.3 contains a remote code execution vulnerability that allows unauthenticated attackers to execute arbitrary code by injecting PHP code through the db_name parameter. Attackers can send a POST request to install/step1.php with malicious PHP code in the db_name parameter, then execute commands via the check.php endpoint using the cmd GET parameter.