Just published.
The newest CVEs, straight from CVE.org and NVD.
-
CVE-2026-71861
The Chainguard SDK is a public Go SDK for integrating with the Chainguard platform. Prior to 0.1.107, auth/aws.VerifyToken trusts the Chainguard-Audience and Chainguard-Identity headers without requiring those headers to be covered by AWS SigV4 SignedHeaders. An attacker with a fresh STS GetCallerIdentity request signed by a victim can rewrite the unsigned binding headers and replay the proof to another relying party, causing the relying party to authenticate the victim’s AWS identity for an attacker-chosen audience and identity. Code that only uses auth/aws.GenerateToken is unaffected because the generator signs the binding headers, and exploitation requires a signed request within its validity window. This issue is fixed in version 0.1.107.
-
CVE-2026-126755
ArangoDB before 3.12.11 contains an authentication bypass vulnerability caused by a parser differential between the raw path authentication gate and the URL-decoded dispatch logic. Attackers can send requests with a %5f-encoded leading underscore to endpoints like /_db/_system/%5fapi/simple/first-example to reach /_api/* actions as superuser without authentication. Exploitation evidence was first observed by the Shadowserver Foundation on 2026-09-16.
-
CVE-2026-109861
CAST Imaging contains a hardcoded authentication token in the "KeyCloakController" class of auth-service.jar. An attacker who gains access to the source code, application package, logs, or configuration files can reuse this token to call auth-service endpoints without authentication. Using the token with the "/api/auth/keycloak/users" endpoint, the attacker can enumerate users across Keycloak realms and retrieve user attributes containing API token hashes, enabling reconnaissance and further exploitation. The vulnerability has only been confirmed in version 3.6.2 but may also affect other versions. The vendor did not provide the affected version range, but assured that the vulnerability is fixed "in newer versions".
-
CVE-2026-132079
Insufficient Session Expiration vulnerability in the Apache Iggy Go SDK. The TCP client method CreatePersonalAccessToken wrote the 32-bit expiry into the upper half of the 64-bit expiry field that the server reads. The server read a nonzero expiry as 2^32 times the requested value. A token created through the Go TCP client therefore stayed valid about four billion times longer than the caller asked for. This issue affects Apache Iggy Go SDK: from 0.5.0 before 0.10.0. Users are recommended to upgrade to version 0.10.0, which fixes the issue. Tokens that an earlier version created keep their long expiry on the server. Delete those tokens and create new ones.
-
CVE-2026-109549
OmniFaces is a utility library for Faces. From 5.2 until 5.4.3, applications using <o:sse> or @Push(type=SSE) with session scope, view scope, or a user attribute can expose pushed messages because org.omnifaces.cdi.push.SseEndpoint#doGet accepts a channel identifier registered by SseChannelManager and PushSessionManager without verifying that the caller's HTTP session owns it. An attacker who obtains the victim's channel identifier from a logged URL, report, Referer header, or shared link can connect from a foreign or absent session and receive messages intended for the victim. Application-scoped channels are intentionally unbound, and this issue permits message receipt but not message injection. This issue is fixed in version 5.4.3.
-
CVE-2026-114924
Vulnerability in the Oracle Application Object Library product of Oracle E-Business Suite (component: Core). Supported versions that are affected are 12.2.3-12.2.15. Easily exploitable vulnerability allows unauthenticated attacker with network access via HTTP to compromise Oracle Application Object Library. Successful attacks of this vulnerability can result in takeover of Oracle Application Object Library. CVSS 3.1 Base Score 9.8 (Confidentiality, Integrity and Availability impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H).
-
CVE-2026-110782
The SimpleSAMLphp SAML2 library is a PHP library for SAML2 related functionality. From 6.2.2 until 6.2.4 and 7.0.0-rc2, HTTPArtifact::receive() in src/Binding/HTTPArtifact.php verifies a signed outer ArtifactResponse but returns an unsigned embedded SAML Response before calling verifyMessageSignature(). The signed path verifies with the artifact issuer's metadata keys without requiring the embedded Response issuer to match that artifact issuer. In a multi-IdP federation, an attacker controlling one trusted IdP can therefore return a validly signed outer ArtifactResponse containing an unsigned Response that claims a different IdP, bypassing the binding's embedded-message authentication. End-to-end impersonation requires a service provider using HTTP-Artifact and a downstream consumer that relies on the binding to authenticate the embedded Response rather than independently requiring a signature. This issue is fixed in versions 6.2.4 and 7.0.0-rc2.
-
CVE-2026-130626
The core JDK HTTP client (a2a-java-sdk-http-client) applies the GHSA-9rhm redirect-neutralization only to the POST path. JdkGetBuilder.get(), JdkGetBuilder.getAsyncSSE(), and JdkDeleteBuilder.delete() call the injected httpClient directly, with no redirect gate. An application that supplies a redirect-capable HttpClient (Redirect.NORMAL/ALWAYS) — which is the SDK-documented way to enable POST redirects — silently follows cross-authority redirects on every credentialed GET/DELETE request. The JDK strips Authorization cross-origin, but forwards the four allowlisted custom credential headers (X-API-Key, API-Key, X-Auth-Token, X-Authentication) to the attacker-controlled authority. This is an incomplete-fix variant of GHSA-9rhm: the fix hardened the default client and added a POST-only two-step opt-in, but never covered GET/DELETE.
-
CVE-2026-130627
The Vert.x HTTP client transport (a2a-java-sdk-http-client-vertx) hand-rolls a POST redirect follower that, when the documented PostBuilder.followRedirects(true) opt-in is enabled, re-issues the redirected request with the full request header map copied verbatim and performs no credential stripping and no same-authority check. A malicious, compromised, or mistyped agent that returns a 30x response with an absolute cross-host Location causes the client to forward the standard Authorization header (Bearer/Basic/OAuth) — plus every other header, including the allowlisted API-key headers — to the attacker-controlled authority. This breaks the project's own cross-transport invariant (SECURITY.md §8, OkHttpPostRedirectHeaderLeakTest) that credentials are stripped on cross-authority redirects even when redirects are followed. The sibling JDK and OkHttp/Android transports strip Authorization; Vert.x uniquely does not. This is a "fix introduced the bug" variant — the follower was added by the GHSA-9rhm hardening PR (3ae82912, refined by a5e76a0b).
-
CVE-2026-57151
GLPI is a free asset and IT management software package. From 11.0.0 until 11.0.10 and 12.0.1, an authenticated technician can modify the two-factor authentication security of a user with higher privileges because the affected operations do not enforce all required authorization checks for the target user. This allows the technician to disable or take control of the higher-privileged user's two-factor authentication protection. This issue is fixed in versions 11.0.10 and 12.0.1.
-
CVE-2026-115563
The Model Context Protocol Python SDK is the official Python SDK for Model Context Protocol servers and clients. From 1.9.1 until 1.30.0 and 2.2.0, OAuth client support in mcp.client.auth can send credentials to an authorization server chosen by an untrusted MCP server. ClientCredentialsOAuthProvider and PrivateKeyJWTOAuthProvider cannot identify the authorization server that owns their credentials. Affected HTTP clients using those providers, OAuthClientProvider, or the deprecated 1.x RFC7523OAuthClientProvider can disclose a client_secret, authorization code, PKCE code_verifier, or signed client assertion. MCP servers, stdio clients, and clients that attach their own tokens or headers are not affected. Interactive OAuthClientProvider use requires a person to start sign-in, while the unattended providers do not. For ClientCredentialsOAuthProvider and PrivateKeyJWTOAuthProvider, callers must configure issuer= after upgrading because omission still follows the MCP server's advertised authorization server, and the deprecated RFC7523OAuthClientProvider must be replaced because it has no issuer option. A stored client registration without an issuer remains unbound after upgrading, so stored OAuth client information must be cleared once to register again or a manually pre-registered record must be updated with its issuer. This issue is fixed in versions 1.30.0 and 2.2.0.
-
CVE-2026-129950
Rowy through 3.0.0 contains an incorrect permission assignment vulnerability in default Firebase Storage rules that grants any user holding a role read and write access to the entire bucket. Authenticated users restricted to a single table or read-only access can use the Firebase SDK directly to list, download, overwrite, and delete any stored files.
-
CVE-2026-74763
Authorization annotations, such as @RequiresRoles and @RequiresPermissions, are not always enforced in some class or interface hierarchies. Annotations declared on an interface were not enforced when using the shiro-guice, shiro-aspectj, shiro-cdi or shiro-jakarta-ee modules. In addition, annotations declared on a class or a superclass were not always enforced when using shiro-guice or shiro-aspectj, allowing an unauthorized caller to invoke methods that should be protected. This issue affects all Apache Shiro versions through 3.0.1, with impact depending on module and where annotations are declared. Mitigation: Upgrade to version 3.0.2 or later, which fixes the issue. The annotation closest to the invoked method now takes precedence. Alternatively, declare authorization annotations directly on each protected implementing method.
-
CVE-2026-67250
A local privilege escalation (LPE) vulnerability in Pebble’s installers could allow a local attacker with low privilege access to place a malicious Dynamic Link Library (DLL) where a privileged executable file will load it.
-
CVE-2026-124278
Authorization Bypass Through User-Controlled Key vulnerability in elixir-grpc grpc_server and grpc allows an attacker to make a handler act on a resource other than the one named in the URL of an HTTP-transcoded route, via a request with a gRPC Content-Type. Only the transcoding clause of GRPC.Server.do_handle_request/4 (grpc_server/lib/grpc/server.ex, and lib/grpc/server.ex before the package split) applies the route's path bindings to the request message. The Cowboy handler picks that clause from the request Content-Type, not from the route. A request to a transcoded URL with application/grpc or application/grpc-web is decoded from the body alone, so every path-bound field takes its value from the body. A gateway, proxy rule or WAF that authorizes on the URL path sees one resource while the handler receives another. This issue affects grpc_server: from 1.0.0 before 1.1.0; grpc: from 0.8.0 before 1.0.0.
-
CVE-2026-128424
A flaw was found in Kroxylicious. The ACL authorizer's rules-file parser unquotes `matching /…/` regular-expression literals using the string unquoting rules, which strips every backslash escape. As a result, an allow rule such as `matching /orders\.v1/` is compiled as the regex `orders.v1`, which also matches unintended names such as `ordersXv1`, and character classes such as `\d` and `\w` lose their special meaning. Allow rules written with regex escapes can therefore apply to more resources than the author intended. Depending on the ACL configuration, this may allow an authenticated Kafka client to access a resource (such as read or write a topic) that should have been denied.
-
CVE-2026-128065
@grpc/grpc-js implements the core functionality of gRPC purely in JavaScript, without a C++ addon. From 1.14.0 through 1.14.1, XdsServer in @grpc/grpc-js-xds can acknowledge Route and VirtualHost resources while failing to apply their RBACPerRoute override configurations. The override parser is not registered under the override message type, override maps are stored under a type URL instead of the configured HTTP filter name, and VirtualHost override maps are not read. When experimental RBAC is enabled and a trusted xDS control plane supplies a broad top-level ALLOW policy with a stricter matching override, affected servers evaluate RPCs under the broad policy and can accept requests that should be denied. This issue is fixed in version 1.14.2.
-
CVE-2026-111456
A missing authorization check has been found in ConsumerResource.setupOwner() function of consumer registration module in the Candlepin package. The registration method checks permission to create a consumer only when the caller is a user. A consumer that authenticates with its identity certificate does not go through that check. A low-privileged network attacker or user who has a valid consumer certificate for one organization can register a new consumer in a different organization by supplying that organization's key. The response includes an identity certificate for the new consumer. With that certificate, the attacker can view the target organization's subscription pools. When the organization is not using simple content access, the attacker can attach that organization's entitlements to the new consumer.
-
CVE-2026-107152
An improper authentication vulnerability has been found in the task console in the smart_proxy_dynflow package. The console accepts any client certificate signed by the capsule's certificate authority. It does not check the list of servers the capsule trusts. A low-privileged network attacker or user must already hold such a certificate. They can open the console and read job details. Those details include passwords when a job was given them. They can also cancel a running task.
-
CVE-2026-126835
The MongoDB Node.js driver does not consistently enforce its permitted-host restriction during OIDC authentication. A bearer token produced by the embedding application's own login callback may therefore be transmitted to a host outside the configured set of permitted hosts, which may result in unintended disclosure of that credential. No database account is required; the party supplying the connection configuration needs only to influence the settings the application passes to the driver.
-
CVE-2026-126644
A link-following flaw was found in the authorized_key module of the ansible.posix collection, residual to the fix for CVE-2026-11837. That fix guards only the final path component (os.chown with follow_symlinks=False and O_NOFOLLOW in _safe_open_write()). When an unprivileged local user replaces their ~/.ssh directory with a symlink to a directory they do not own, and an operator subsequently runs the authorized_key task as root (a common provisioning pattern), the module resolves its path operations through the symlink: atomic_move() writes the temporary file in the symlink target and renames it to <target>/authorized_keys, creating or overwriting a file in the attacker-chosen directory, and os.chmod(sshdir, 0o700) follows the symlink and changes the target directory's permissions. The written file is chowned to the unprivileged user, so this is not a privilege escalation, but it allows an attacker to create or overwrite files (for example in /etc/apt/apt.conf.d/, which APT reads regardless of file ownership) to cause integrity and availability impacts.
The newest published CVE records from CVE.org and NVD, refreshed every few minutes, with severity, exploit odds and fix status.