| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| ActivityPub is a multi-tenant ActivityPub server for Ghost, built with Fedify. From 1.0.0 until 1.2.6, the `Undo` handler allows a remote federated server to cause the service to act on a `follow` or `repost` as though the activity came from a different account. This permits removal of an existing `follow` relationship or `repost` belonging to another account without that account's involvement. No private data is exposed and content is not modified; the affected relationships can be re-established. Services proxying to Ghost's hosted ActivityPub server have already been updated, while self-hosted services require an updated image and redeployment. This issue is fixed in version 1.2.6. |
| A security vulnerability has been detected in DAMO Academy VideoLLaMA2 up to 1.1. Affected by this issue is the function load_pretrained_model of the file videollama2/model/__init__.py of the component Model Loading. The manipulation of the argument _name_or_path leads to improper validation of integrity check value. Remote exploitation of the attack is possible. The exploit has been disclosed publicly and may be used. The project was informed of the problem early through an issue report but has not responded yet. |
| Munshi (by Petpooja) before 2026-06-30 contained a payment bypass vulnerability in its subscription payment webhook at POST /api/payment/order/webhook, which lacks webhook authentication, cryptographic signature verification, source validation, and independent server-side confirmation with the payment provider. A remote, unauthenticated attacker can submit a forged payment-success notification referencing an order obtained from the payment-status redirect URL, causing the backend to mark the subscription as paid/active and granting paid access without an actual successful payment. |
| BIND may accept incorrect child-zone NSEC3 records as valid, which could allow an attacker to forge authenticated NXDOMAIN responses.
This issue affects BIND 9 versions 9.18.0 through 9.18.50, 9.20.0 through 9.20.24, 9.21.0 through 9.21.23, 9.11.3-S1 through 9.18.50-S1, and 9.20.9-S1 through 9.20.24-S1. |
| It is possible for an attacker's zone to respond to a query with an RRSIG that has a smaller number of labels than the zone in which the RRSIG is contained. This causes `named` to produce a wildcard name for a zone that is shorter than the attacker's zone, which can result in cache poisoning. For this attack to have any effect, the resolver under attack must have set `synth-from-dnssec yes;` (which is the default).
This issue affects BIND 9 versions 9.11.0 through 9.18.50, 9.20.0 through 9.20.24, 9.21.0 through 9.21.23, 9.11.3-S1 through 9.18.50-S1, and 9.20.9-S1 through 9.20.24-S1. |
| A vulnerability was detected in Yoosee GM8135S Camera 21.00.01.36. Affected is an unknown function of the component Secure Boot. Performing a manipulation results in improper verification of cryptographic signature. The attack needs to be approached locally. The exploit is now public and may be used. The vendor was contacted early about this disclosure but did not respond in any way. |
| http4k from 4.9.10.0 before 6.60.0.0 contains a missing integrity check vulnerability because DigestEncoder computes the auth-int digest without hashing the request entity body. On-path attackers on non-TLS channels can rewrite the body of authenticated requests while keeping the Authorization header valid, executing state-changing actions as the victim. |
| The Mollie webhook endpoint accepts two independent, attacker-controlled parameters and never verifies that they belong together:
* commerceTransactionHash — selects which Craft Commerce transaction is completed
* id — selects which Mollie payment is queried for its status
An unauthenticated attacker holding any paid Mollie payment ID from the merchant’s account can pair it with any transaction hash.
Commerce queries Mollie, sees status: paid, and marks the target transaction successful for that transaction’s full
amount, regardless of how much was actually paid. |
| Postiz signs its session, invitation and password reset tokens with one shared secret and without any marker of what a token is for. A route that refreshes the credentials of a browser-extension channel signed data taken from the request itself and stored the result where the requesting user can read it back. A user could therefore obtain a validly signed token containing any claims they choose, which the application accepts as a session, as an organization invitation with a chosen role, as a password reset or as an enterprise provisioning token.
A second, narrower route that creates team invitations signs extra fields from the request body in the same way. |
| 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. |
| 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. |
| A flaw was found in dnsmasq. When following a CNAME in a DNS response, the resolver can cache records for the CNAME target without checking that those records belong to the responding server's authorized zone. In split-DNS deployments that send some domains to a specific upstream, an attacker who controls that upstream can inject cache entries for unrelated names. Later clients may then receive attacker-chosen addresses for those names. |
| 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. |
| Zuke is a code-first, strongly typed build automation system for Deno. Prior to 1.10.0, the @zuke/gh githubWorkflow(...) gate correlates a dispatched workflow run using only a display-title marker that is visible in the workflow run list and can be copied by another dispatcher. In configurations where the external workflow repository grants workflow_dispatch permission to a party not equally trusted by the releasing repository, an attacker can dispatch a run with the marker and cause the gate to adopt its conclusion without verifying the event, ref, or creation window. A passing substituted run can authorize an unvalidated release, while a failed or cancelled run can block a valid release, and the persisted adopted run ID preserves the substitution across build resume. This issue is fixed in version 1.10.0. |
| 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. |
| 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. |
| A Kafka broker using SASL/OAUTHBEARER with the default validator (sasl.oauthbearer.jwt.validator.class=DefaultJwtValidator) selects its implementation from configuration. When sasl.oauthbearer.jwks.endpoint.url is set, the validator verifies the JWT signature. When that URL is unset (the default) and no verification key resolver is supplied, DefaultJwtValidator silently falls back to ClientJwtValidator, which does not verify the JWT signature, algorithm, or issuer - and the broker starts with no warning or error. This is the residual of CVE-2026-33557, whose fix restored signature verification only on the path where the JWKS endpoint URL is set.
Impact:
A broker in this state accepts any structurally valid JWT, including "alg":"none" tokens. An unauthenticated remote attacker who can reach the listener can forge a token asserting any principal - including configured superusers - and gain full access to the cluster. Only brokers that enable SASL/OAUTHBEARER validation without a JWKS endpoint URL (and without a custom verification key resolver) are affected; brokers that set the URL verify signatures and are not affected. Unlike CVE-2026-110952 (issuer/audience not enforced by default), here no signature is verified at all.
Mitigation (all affected versions, no upgrade required):
Set sasl.oauthbearer.jwks.endpoint.url for every listener/mechanism that uses OAuthBearerValidatorCallbackHandler, or configure a signature-verifying sasl.oauthbearer.jwt.validator.class.
Resolution:
From Apache Kafka 4.2.2, 4.3.1, and 4.4.0: a broker that would fall back to the non-verifying ClientJwtValidator fails to start with a ConfigException instead of running without signature verification (secure by default). |
| Pipecat is an open-source Python framework for building real-time voice and multimodal conversational agents. From 0.0.98 until 1.6.0, the pipecat-ai package's built-in development runner exposes unauthenticated endpoints including /start, /daily, and /daily-dialin-webhook that can create bot sessions and Daily rooms. The runner binds to localhost by default and is intended only for local development, so exploitation requires an operator to expose it to untrusted traffic through --host 0.0.0.0, a published port, or a tunnel without access controls. An unauthenticated caller can consume local resources and third-party quotas and, depending on configuration, receive a room URL and join token. Version 1.6.0 adds a prominent development-runner banner and documentation that clarifies the unauthenticated boundary and directs production deployments to an authenticated dispatcher or managed runtime, while the endpoint behavior remains by design. This issue is fixed in version 1.6.0. |
| In `github.com/eclipse-keypont/crypto11/v2` versions `>= 2.0.0-rc3, <= 2.0.0-rc8`, the ML-KEM key-pair lookup functions (`FindMLKEMKeyPair`, `FindMLKEMKeyPairs`, `FindMLKEMKeyPairsWithAttributes`) locate the public half of an ML-KEM key pair by `CKA_ID`/`CKA_LABEL` and never verify it against the private key. In addition, the "ID only" fallback search runs even when the private key has no `CKA_ID`, so the search template reduces to the public-key class and the ML-KEM key type and the first ML-KEM public key on the token is returned. `Encapsulate()` on the resulting key pair encrypts the fresh shared secret to that key, so its owner — not the holder of the private key the caller looked up — can recover the secret. An attacker who can open an unauthenticated session on the token and plant a public `CKK_ML_KEM` object carrying the victim's `CKA_ID`/`CKA_LABEL` receives the shared secrets for encapsulations performed through the substituted key pair, compromising the confidentiality of those secrets and of everything protected by keys derived from them. The v1 module paths have no ML-KEM support and are not affected. |
| The BitPay Gateway for WooCommerce WordPress plugin through 4.1.3 does not bind the payment provider's transaction confirmation to the order it completes, nor verify the confirmed amount, allowing an unauthenticated attacker who holds a single genuine low-value transaction to mark arbitrary higher-value orders as paid. |
