A command injection vulnerability exists in rpm when processing tarball paths or filenames via the `rpmbuild -t*` command. An attacker who can control these inputs in automated build or CI workflows may execute arbitrary commands with the build user's privileges. This could lead to information disclosure or disruption of the build environment.
A vulnerability in libsolv, a dependency-resolution library used by RPM-based package managers like dnf and zypper, allows an out-of-bounds memory write when processing specially crafted or corrupted .solv cache files. This flaw can cause the affected package management tools to crash, resulting in a denial of service. The vulnerability does not enable arbitrary code execution because the out-of-bounds write stores a fixed, non-attacker-controlled value.
In Twisted versions 25.5.0 and earlier, the wildcardToRegexp() function in twisted/mail/imap4.py improperly handles certain IMAP pattern characters by passing them directly to Python's regular expression compiler. This allows an authenticated client to craft patterns that cause catastrophic backtracking, leading to a denial of service by suspending all server input and output during the match. No fixed release is currently available.
A vulnerability in m17n-lib can cause a denial of service due to a null pointer dereference during library initialization failure. This occurs under specific error conditions like resource exhaustion or database corruption when an application opens an input method. The flaw affects multiple specific versions of m17n-lib and results in application crashes without impacting confidentiality or integrity.
Red Hat has issued a security advisory for its Hardened Images RPMs, specifically updating multiple Rust-related packages to address three security vulnerabilities identified as CVE-2026-109349, CVE-2026-109350, and CVE-2026-109351. The update includes various Rust components such as cargo, clippy, rust-analyzer, rustfmt, and several rust-std-static packages across multiple architectures. No specific affected versions are detailed, and no explicit fix version is provided. The advisory is classified as medium severity.
Multiple arbitrary code execution vulnerabilities have been identified in the xorg-x11-server component of the X.Org open-source implementation of the X Window System. These include use-after-free, heap-based buffer overflow, integer truncation, buffer overflow, and out-of-bounds write issues. The vulnerabilities affect various versions of xorg-x11-server and have been rated with high severity by Red Hat Product Security. An important security update is available for Red Hat Enterprise Linux 9 and related distributions to address these issues.
CVE-2026-111866 is a medium severity SQL injection vulnerability in the nemoTyrant manong product affecting the function cache/del_cache/arr2sql in func.php. The vulnerability allows remote attackers to manipulate SQL queries. The product uses a rolling release model, and no specific affected or fixed versions are identified. The related GitHub issue was closed as completed, but no clear fix commit is known.
Online Robust Mahalanobis is a statistical method for anomaly detection in robotic joint residuals. It measures how far a new sample deviates from a healthy baseline distribution using a robust Mahalanobis distance metric that is resistant to outliers. The method updates estimates of location and scatter online using averaged stochastic gradient descent, enabling real-time anomaly scoring. It is designed to detect faults such as payload changes in robotic arms by flagging samples that fall outside a learned healthy range. The approach improves on classical Mahalanobis distance by using robust estimators (geometric median and median covariation matrix) to avoid masking effects caused by outliers. This technique is presented as a research and utility module rather than a vulnerability or active threat.
This content discusses the use of Lynis, a security auditing and hardening tool for Linux and UNIX-based systems, to evaluate the security posture of various Linux distributions using their default configurations. It presents a comparative overview of Lynis Hardening Index scores across multiple distributions, indicating the relative security measures in place by default. The information is intended as a basic comparison and not a definitive security assessment.
Coder versions from 2.10.0 up to but not including 2.34.8 contain an unauthenticated debug HTTP server listening on localhost (127.0.0.1) port 2113. The /debug/manifest endpoint exposes the full agent manifest, including environment variables provided by templates. This allows a lower-privileged Unix user or process within the same workspace to access sensitive credentials such as cloud keys or registry tokens. The vulnerability has a medium severity rating with a CVSS score of 3.3.
Coder versions before 2.29.20, 2.33.0 before 2.33.12, 2.34.0 before 2.34.8, and 2.35.0 before 2.35.4 have an improper hostname validation vulnerability. The flaw is in the GetWorkspaceProxyByHostname function, where a SIMILAR TO pattern performs a prefix match, causing any domain that is a string prefix of a workspace-proxy access URL to be trusted as a proxy. This allows attackers controlling such a domain to trick authenticated users into visiting a crafted URL, capture their application_connect API key, and replay it to access the victim's workspace applications.
Coder versions from 2.0.0 up to but not including 2.34.8 have an authorization vulnerability in the acl_allow rule of coderd/rbac/policy.rego. This flaw allows authenticated users who have been removed from an organization to retain access to organization-scoped resources such as templates and workspaces via the API. The vulnerability arises because the acl_allow rule does not verify organization membership for per-user ACL entries, and removal from the organization does not clear those entries.
Coder versions from 2.30.0 up to but not including 2.34.8 have an authorization bypass vulnerability in the provisioner daemon's DownloadFile gRPC handler. This flaw allows authenticated provisioner daemons with an organization-scoped key and knowledge of a file UUID to download cached Terraform module archives belonging to other organizations. The exposure of these module archives can reveal sensitive infrastructure details or embedded secrets. The vulnerability has a medium severity rating with a CVSS score of 5.3.
Coder versions before 2.29.17, 2.30.0 before 2.32.7, 2.33.0 before 2.33.8, and 2.34.0 before 2.34.2 contain a stored HTML injection vulnerability. This occurs because the ApplicationName and LogoURL appearance settings are rendered into HTML output without sufficient escaping. Attackers with the Owner role can inject HTML markup that is displayed in the Coder dashboard and in SMTP notification emails sent to other users. The vulnerability has a medium severity with a CVSS score of 3.4.
Coder versions from 2.0.0 up to but not including 2.34.8 contain an open redirect vulnerability in the handleAPIKeySmuggling function. This flaw allows attackers to craft URLs on trusted workspace app subdomains with paths starting with double slashes, causing scheme-relative redirects to attacker-controlled sites. Such redirects can be exploited for phishing attacks by misleading users to malicious external sites.
Coder versions from 2.26.0 up to but not including 2.34.8 have an authorization flaw in the GET /api/v2/workspaces/{id}/acl endpoint. This flaw allows authenticated users with workspace use access to retrieve group members' emails, login types, status, and timestamps by exploiting insufficient authorization checks. The vulnerability does not impact integrity or availability but exposes some sensitive information.
Admidio versions from 5.0.0 up to but not including 5.0.16 contain an authorization bypass vulnerability. This flaw allows authenticated users with list view rights to export member lists as PDF files despite export restrictions. The issue arises because the groups_roles_export check in the relevant PHP module omits the 'pdfl' mode, which is used for PDF exports.
Admidio versions from 5.0.0 up to but not including 5.0.16 contain an insecure direct object reference (IDOR) vulnerability in the ForumService::saveTopic() function. This flaw allows authenticated users to modify forum topics belonging to other users by manipulating the topic_uuid parameter in their topic_save requests. Exploitation requires authentication and reuse of the attacker's own CSRF token to overwrite the victim's topic title and first post.
Admidio versions from 5.0.11 up to but not including 5.0.16 contain a broken access control vulnerability in the inventory module. This flaw allows any authenticated user to bypass the inventory_module_enabled restriction and access the DataTables JSON feed, exposing all inventory items and keeper names even when the module is disabled.
Admidio versions prior to 5.0.16 have an information disclosure vulnerability that allows unauthenticated attackers to access role-restricted weblinks. This occurs because the rss/links.php script does not apply the category_view role-visibility filter, exposing link names, URLs, descriptions, categories, and creator names from restricted link categories via the organization's RSS feed.