| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| In the Linux kernel, the following vulnerability has been resolved:
vsock/vmci: validate packet source for connected sockets
vmci_transport_recv_stream_cb() looks up sockets first by the full source
and destination tuple, then by destination only in the bound table. The
fallback can select a non-listening socket without checking whether the
packet came from its stored peer.
This was reproduced with two VMCI contexts. A RST from the context not
stored in a TCP_SYN_SENT socket reset that socket after it was selected by
the destination-only lookup.
VMCI can process notification packets in bottom-half context when the
socket is not owned by user context, or defer packets to a workqueue. Use
vsock_check_source() after taking the socket lock in the bottom-half path,
and recheck after lock_sock() in the workqueue path. Listening sockets
continue to accept packets from any source.
Reply with a RST addressed from the received packet before dropping a
source that fails validation. This preserves the existing reset behavior
for bound non-listening and concurrently closed sockets without directing
the reset to a connected socket's stored peer. |
| A vulnerability was identified in rathena FluxCP up to 98cd4b8fd83228c11b36fb3d097b8e0f7c105a8d. This issue affects the function verifyiprange of the file lib/Flux/PaymentNotifyRequest.php of the component PayPal IPN Handler. The manipulation of the argument X-Real-IP leads to origin validation error. The attack is possible to be carried out remotely. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure but did not respond in any way. |
| vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowed_origins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on localhost, an internal network, or a VPN, call endpoints including GET /v1/models and POST /v1/chat/completions, consume GPU resources, and read responses. Cookie-based cross-site request forgery does not apply because credentials are disabled by default, and deployments requiring bearer authentication are exposed only when the caller already has a usable token. This issue is fixed in version 0.32.0. |
| Python Social Auth is a social authentication and registration mechanism. Prior to 6.2.0, OpenIdConnectAuth in social_core/backends/open_id_connect.py uses get_and_store_nonce to place the OpenID Connect nonce in a shared OpenIdConnectAssociation table, and get_nonce and validate_claims verify only that the nonce exists and has not expired rather than binding it to the browser session and OAuth state that initiated the login. An attacker who obtains a victim's unused authorization code can start a separate login to obtain valid state for the attacker's browser, submit the victim's code with that state, and pass nonce validation because the victim nonce is globally available. With DEFAULT_USE_PKCE disabled in the default OpenIdConnectAuth configuration, the attacker can be logged in as the victim. This issue is fixed in version 6.2.0. |
| Transmission 4.0.0 before 4.1.3 contains a permissive CORS vulnerability in handle_request() in rpc-server.cc that exposes the X-Transmission-Session-Id anti-CSRF token to any web origin. Malicious web pages can read the token from the 409 handshake response and replay it to issue RPC calls, enabling script-torrent-done execution, download-dir changes, or torrent and data deletion. |
| Ladybird is a truly independent web browser. Prior to the c0d68bddf3e3616e5173835f0f4c155ee9700d91 commit, CSSStyleSheet.cssRules, insertRule(), and deleteRule() in Libraries/LibWeb/CSS/CSSStyleSheet.cpp did not enforce the stylesheet origin-clean flag with the required SecurityError, even though cross-origin imported stylesheets were correctly marked non-origin-clean. A malicious page could read parsed rules from a stylesheet loaded without CORS and could insert or delete rules in the victim stylesheet, bypassing the Same-Origin Policy. Reading could expose selectors, generated content, feature markers, or other sensitive rule text, while mutation could alter page visibility or layout. The issue does not provide memory corruption, code execution, or persistence after the attacker page closes. This vulnerability is fixed by commit c0d68bddf3e3616e5173835f0f4c155ee9700d91. |
| Juspay Hyperswitch before 1.128.0 contains an authentication bypass vulnerability that allows unauthenticated attackers to forge payment status updates because the Worldpayxml connector's verify_webhook_source always returns true without enforced mTLS. Attackers knowing the merchant_id and Worldpay orderCode can POST unsigned XML with lastEvent SETTLED to the worldpayxml webhook endpoint, marking unpaid payments as charged or triggering refund, dispute and payout events. |
| WIT SynApps Runtime through 1.4.11 contains a browser-mediated cross-origin request forgery vulnerability that allows attackers to supply an arbitrary target address during controller onboarding, causing the victim's browser to issue discovery requests to attacker-specified hosts at paths such as /easy/GUI/Skins/Default/SiteLogo and /easy/GUI/Skins/Default/Identity. Attackers can leverage the victim's browser reachability to probe loopback addresses, local hostnames, or internal-only network hosts and initiate the enrollment flow against REDY instances, crossing IT/OT segmentation boundaries through a sanctioned application path. |
| LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From 1.87.0 until 1.100.4, 1.101.3, 1.102.2, 1.103.1, and 1.104.0-rc.2, the proxy reuses one cryptographic context for at-rest secrets and session tokens, allowing an authenticated internal user to have crafted metadata containing a forged administrator credential encrypted and returned by the API-key flow. Replaying that ciphertext as a bearer token causes the proxy to trust the forged proxy administrator identity, permitting administrative access and arbitrary commands through the MCP stdio endpoint. Versions 1.87.0 through 1.90.x require EXPERIMENTAL_UI_LOGIN to be enabled, while versions after 1.91.0 are exploitable in the default configuration. This issue was fixed in LiteLLM 1.100.4, 1.101.3, 1.102.2, 1.103.1, and 1.104.0-rc.2. |
| Apache Pulsar's MultiRolesTokenAuthorizationProvider authorizes a client with all roles listed in a claim of its JSON Web Token (JWT). The provider is intended for use with token authentication, which verifies the token before authorization. The provider did not verify the token itself: it parsed the token's claims with the signature removed and trusted that authentication had already checked it.
When authentication and authorization are enabled with a non-empty anonymousUserRole setting, some HTTP requests whose credentials fail authentication are not rejected, but are handled as requests of the anonymous user role. For the broker's HTTP endpoints, such as the admin REST API, this fallback applies to requests that do not name an authentication method and requires strictAuthMethod to be false (the default) where that setting is available. For the WebSocket API, failed credentials can also fall back to the anonymous user role; this path does not enforce strictAuthMethod. The authentication data kept for such a request still contained the request's Authorization header. MultiRolesTokenAuthorizationProvider then read the bearer token from that header without verifying its signature and authorized the request with the roles in the token's role claim.
As a result, in deployments that use MultiRolesTokenAuthorizationProvider together with a non-empty anonymousUserRole and that permit this fallback, an unauthenticated client can send an HTTP request, for example to the admin REST API or the WebSocket API, with a forged or unsigned token that lists arbitrary roles, and be authorized with those roles. The impact can include super-user access when the token lists a role configured in superUserRoles, allowing the client to read and change tenants, namespaces, topics, permissions and broker configuration. No valid credentials are required.
Deployments are not affected if they use the default authorization provider (PulsarAuthorizationProvider) or another authorization provider, or if anonymousUserRole is not set (the default). Requests that authenticate successfully with a valid token are not affected, because the token that MultiRolesTokenAuthorizationProvider reads is the token that authentication verified. Connections that use the Pulsar binary protocol are not affected by this issue, because the broker does not keep client authentication data for connections that it assigns to the anonymous user role.
In fixed versions, MultiRolesTokenAuthorizationProvider obtains roles only from tokens that the configured token authentication provider has verified, using its configured keys and validation settings. Broker HTTP requests that are assigned to the anonymous user role are authorized only as that role, and invalid tokens cannot supply additional roles to WebSocket requests. The broker fails to start if MultiRolesTokenAuthorizationProvider is configured with authorization enabled but without a token authentication provider.
This issue affects Apache Pulsar versions from 2.8.1 before 4.0.14 and from 4.1.0 before 4.2.5.
- 4.0.x version users should upgrade to at least 4.0.14.
- 4.1.x and 4.2.x version users should upgrade to at least 4.2.5.
Users operating versions prior to those listed above should upgrade to the aforementioned patched versions or newer versions.
The currently supported Apache Pulsar release lines are 4.0.x and 5.0.x. The supported versions schedule is available at https://pulsar.apache.org/contribute/release-policy/.
Users who cannot upgrade immediately can remove the anonymousUserRole setting from the broker and WebSocket service configurations, or restrict network access to the HTTP endpoints of these services to trusted clients. |
| GitLab has remediated an issue in GitLab EE affecting all versions from 17.8 before 19.3.6, 19.4 before 19.4.3, and 19.5 before 19.5.1 that under certain conditions could have allowed an authenticated user to access paid AI features without a valid subscription due to insufficient server-side validation in billing eligibility checks. |
| PraisonAI is a multi-agent teams system. From version 4.5.2 to before version 4.6.59, praisonai.bots.DiscordApproval approves a pending dangerous tool call when it sees any later non-bot message in the configured Discord channel whose text is classified as approval, such as yes. As a result, any user who can post in the configured approval channel can approve a pending high-risk tool call by sending yes after the approval message appears. the same PoV also shows that the Slack and Telegram messaging approval backends have no explicit approver allowlist parameter, but the primary report-grade issue is the Discord backend's unthreaded channel cross-talk: the approving message does not need to be a reply or otherwise request-bound. This issue has been patched in version 4.6.59. |
| phpMyAdmin through 5.2.3 contains an authentication bypass vulnerability in CustomServer::parseAndValidateAssertionResponse() that never verifies WebAuthn assertion signatures against stored public keys. Attackers holding a victim's MySQL credentials can POST a crafted webauthn_request_response with matching challenge, origin, and rpIdHash to bypass the second factor and gain database privileges. |
| Improper verification of cryptographic signature in Microsoft Edge (Chromium-based) allows an unauthorized attacker to perform spoofing over a network. |
| In the Linux kernel, the following vulnerability has been resolved:
smb: client: reject userspace cifs.idmap descriptions
cifs.idmap key descriptions carry authority-bearing fields (owner and
group SIDs and uid/gid values in "os:"/"gs:"/"oi:"/"gi:" form) that the
cifs.idmap upcall helper treats as kernel-originating inputs. Unlike
its sibling cifs.spnego, the cifs.idmap key type has no vet_description
hook, so userspace can create keys of this type through
request_key(2)/add_key(2) and supply those fields without CIFS origin.
A request_key(2) call with a non-NULL callout then drives a root
usermodehelper upcall (/sbin/request-key -> cifs.idmap) that consumes
the unvetted description in root context.
Only accept cifs.idmap descriptions while CIFS is using its private
root_cred to request the key. id_to_sid()/sid_to_id() already run
under override_creds(root_cred), so the kernel-originated path is
unaffected.
This mirrors commit 3da1fdf4efbc ("smb: client: reject userspace
cifs.spnego descriptions"), which applied the same restriction to
cifs.spnego. |
| Echo is a Go web framework. From 4.0.0 until 5.4.0, the Proxy middleware can preserve an attacker-supplied X-Real-IP header and forward it to an upstream service when Echo.IPExtractor is not configured. A remote client can therefore make upstream services receive a spoofed client address, undermining IP-based access controls, rate limiting, audit logging, or geographic restrictions that trust Echo's forwarded identity. This issue is fixed in Echo 5.4.0. |
| Use of a less trusted source for the envelope sender in the external-account download of Progressive Robot hMailServer 6.0.0 through 6.3.6 allows a remote unauthenticated attacker to have mail they write in the name of a domain the server hosts sent on by the server with an envelope sender at that domain, so that it passes the domain's SPF and DMARC at other receivers. A message the server downloads from an external POP3 or IMAP account has no SMTP envelope of its own, and the server took the address in its From header as its envelope sender, whatever that address was. A forward or redirect of the downloaded message to an outside address then left with that address as its envelope sender, which the Sender Rewriting Scheme leaves alone for a hosted domain, so that the receiver's SPF check passed for the forged domain, aligned even under strict alignment, with no non-default setting, host name record or DMARC policy needed. An attacker needs to know a mailbox elsewhere that an account here collects and forwards to an outside address; an account holder who adds an external account of their own can collect from a server of their own, which matters where SmtpAuthenticatedSenderCheck holds signed-in senders to their own addresses. |
| rclone from 1.55.0 before 1.75.2 contains a signature verification bypass in stable selfupdate that trusts checksums from unsigned bytes preceding the PGP-clearsigned SHA256SUMS block. Attackers controlling the release distribution path can prepend an unsigned checksum line matching a malicious archive, causing selfupdate to install an attacker-controlled rclone executable. |
| Twisted is an event-based framework for internet applications, supporting Python 3.6+. Prior to 26.10.0rc1, twisted.names.dns.DNSDatagramProtocol accepts a UDP DNS response for a pending query when its 16-bit transaction ID matches but does not verify that the response came from the queried host and port. An attacker able to deliver a packet to the resolver's listening port with the correct transaction ID can inject a forged answer. The standard twisted.names.client.Resolver uses a new random source port for each query, increasing off-path attack complexity, while applications that reuse a DNSDatagramProtocol instance expose only the transaction-ID search space after the first query. A successful forgery can redirect hostname resolution and enable mail interception, server-side request forgery, or man-in-the-middle positioning. This issue is fixed in version 26.10.0rc1. |
| AnkiMCP is an Anki addon that implements an MCP server, enabling AI assistants to interact with the Anki flashcard application. Prior to version 0.21.0, `anki_mcp_server/mcp_server.py` globally disables FastMCP DNS-rebinding protection through `TransportSecuritySettings` with `enable_dns_rebinding_protection=False`, including for the default local Streamable HTTP listener. The defaults in `anki_mcp_server/config.py` enable `http_enabled` on `127.0.0.1:3141`, with an empty `http_path` selecting the root endpoint. No bearer token, API key, session secret, strict `Host` allowlist or strict `Origin` allowlist is enforced before `initialize`, `tools/list` and `tools/call` reach MCP dispatch. A victim who visits a malicious web page while Anki and the listener are running can have browser requests rebound to loopback, allowing the attacker to `initialize` MCP, enumerate tools, read the local collection, and perform enabled non-destructive note/deck creation or updates. The attacker needs neither LAN position nor a local process. Available operations depend on enabled tool filters; destructive tools are hidden unless `enabled_destructive_tools` is opted into, limiting deletion-level impact but not the exposed read/write operations. The reported harness confirms real MCP tool dispatch with substituted Anki bridge responses rather than demonstrating destructive changes to a real collection. Disabling protection for tunnels/proxies also disables it for the normal loopback configuration. This issue is fixed in version 0.21.0. |
