Metabase is an open-source business intelligence and embedded analytics tool. Prior to 1.58.15, 1.59.12, 1.60.6.3, and 1.61.1.4, Metabase instances with an H2 database connection, including the default sample database, deserialize arbitrary Java objects returned in H2 native query result columns of type OTHER without validation, allowing an authenticated user who can run native H2 queries to execute code on the Metabase server. This issue is fixed in versions 1.58.15, 1.59.12, 1.60.6.3, and 1.61.1.4.

Metabase is an open-source business intelligence and embedded analytics tool. From 1.55.0 until 1.58.15.1, 1.59.12, 1.60.6.3, and 1.61.2, Metabase did not validate unsafe H2 connection properties on one database-creation code path, allowing an authenticated administrator to register a crafted H2 database connection and execute arbitrary Java code on the Metabase server. This issue is fixed in versions 1.58.15.1, 1.59.12, 1.60.6.3, and 1.61.2.

Mastodon is a free, open-source social network server based on ActivityPub. Prior to 4.4.19 and from 4.5.0 until 4.5.12, Mastodon's app/models/concerns/user/ldap_authenticable.rb mutates OpenSSL::SSL::SSLContext::DEFAULT_PARAMS when LDAP authentication uses LDAP_TLS_NO_VERIFY=true, disabling SSL and TLS certificate verification globally for requests made by puma web processes while sidekiq background jobs remain unaffected. This issue is fixed in versions 4.4.19 and 4.5.12.

LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.84.0, LiteLLM's MCP Streamable HTTP endpoint allowed an unauthenticated attacker to use a fabricated Authorization header to trigger an OAuth2 passthrough fallback path that replaced failed LiteLLM key validation with an empty UserAPIKeyAuth() object, allowing requests to reach MCP tooling without a valid LiteLLM key. This issue is fixed in version 1.84.0.

LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.82.0-stable, LiteLLM's Custom Code Guardrails production create and update paths did not apply the same sandboxing and validation used by the test endpoint, allowing a privileged user with access to create or update guardrails to submit custom Python code that executed in the LiteLLM proxy environment and could expose secrets available to the process. This issue is fixed in version 1.82.0-stable.

LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.7-stable, LiteLLM Skills archive extraction did not sufficiently validate file paths from uploaded skill ZIP archives, allowing an authenticated user with access to LiteLLM LLM API routes or a key whose allowed_routes includes /v1/skills, anthropic_routes, or llm_api_routes to upload a crafted skill archive containing path traversal entries that could be written outside the intended extraction or staging directory. This issue is fixed in version 1.83.7-stable.

A vulnerability has been found in D-Link DIR-605L 2.13B01. This affects the function formAdvFirewall of the file /goform/formAdvFirewall of the component POST Request Handler. Such manipulation of the argument curTime leads to buffer overflow. The attack may be launched remotely. The exploit has been disclosed to the public and may be used. This vulnerability only affects products that are no longer supported by the maintainer.

LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.10-stable, LiteLLM's /health/test_connection endpoint resolved request-supplied environment and OIDC file references in litellm_params, allowing a proxy administrator or another privileged caller with permission to test model connections to read files from the local filesystem via an oidc/file/ reference. This issue is fixed in version 1.83.10-stable.

etcd is a distributed key-value store for the data of a distributed system. Prior to 3.5.32 and 3.6.13, when etcd is configured with --listen-client-http-urls to split HTTP and gRPC client endpoints onto separate listeners, the --client-crl-file Certificate Revocation List is not enforced on the gRPC listener, allowing a client with a revoked certificate to authenticate successfully over gRPC. This issue is fixed in versions 3.5.32 and 3.6.13.

Ghost is a Node.js content management system. From 6.27.0 before 6.44.0, Ghost's public donation checkout flow allowed an unauthenticated attacker to control donation checkout metadata and obtain full paid gift memberships for a minimal payment without exposing customer or member data or stealing money from a site or its members. This issue is fixed in version 6.44.0.

A flaw has been found in D-Link DIR-605L 2.13B01. Affected by this issue is the function formSetMACFilter of the file /goform/formSetMACFilter of the component POST Request Handler. This manipulation of the argument curTime causes buffer overflow. The attack may be initiated remotely. The exploit has been published and may be used. This vulnerability only affects products that are no longer supported by the maintainer.

SiYuan before v3.8.0 interpolates secret placeholders into the destination URL parameter of the http_request MCP tool, allowing attackers to exfiltrate stored secrets. An MCP client can craft a request with an attacker-controlled URL containing secret placeholders to send plaintext secret values to any public host without confirmation.

AVideo through commit 9c39d8c8 contains an authentication bypass vulnerability where deduplicateByEncoderQueueId() returns video_id_hash credentials for any video by encoder_queue_id without ownership verification, and useVideoHashOrLogin() converts this hash into passwordless login as the video owner. Attackers with upload permission can retrieve an administrator's video_id_hash by omitting the videos_id parameter, then use that hash in an unauthenticated request to gain administrative session access and modify system configuration.

Composio SDK before 0.2.32-beta.283 contains a path validation bypass vulnerability that allows attackers to read and exfiltrate sensitive files by exploiting a missing assertSafeFileUploadPath check in the readFileFromDisk function within tool-file-uploads.ts. Attackers can exploit prompt injection to manipulate file_uploadable parameters to reference sensitive paths such as SSH private keys, causing the CLI to upload credential files to attacker-controlled storage.

Gradio before 6.20.0 contains an open redirect and server-side request forgery vulnerability that allows attackers to redirect users to arbitrary URLs or perform client-side SSRF by supplying unvalidated HTTP/HTTPS URLs to the file_fetch() function in the /gradio_api/file= endpoint. Attackers can craft a malicious FileData response targeting internal endpoints such as cloud metadata services to retrieve sensitive credentials including EC2 IAM role credentials.

Gumroad before 2026.07.06.2 contains a broken access control vulnerability in the PurchasesController that allows authenticated sellers to manipulate purchase access for other sellers' products by sending PUT requests to the revoke_access and undo_revoke_access actions without seller ownership validation. Attackers can modify the is_access_revoked status on arbitrary purchases to unauthorized revoke or restore buyer access to products they do not own.

Midscene Bridge Server through 1.10.3, fixed in commit 86f4118, contains a missing authentication and CORS misconfiguration vulnerability that allows unauthenticated remote attackers to hijack active bridge sessions by opening a cross-origin WebSocket connection to the local Socket.IO server, which performs no Origin header validation and requires no authentication token. Attackers can connect from any web page visited by the victim to seize the single-client slot, intercept and inject automation commands, exfiltrate command-payload data, or unconditionally terminate the server by supplying the MIDSCENE_BRIDGE_SIGNAL_KILL query parameter.

rpcx through 1.9.3, fixed in commit 047aec1, contains a denial-of-service vulnerability in protocol.Message.Decode (protocol/message.go). When a message has the compression flag set, the payload is gzip-decompressed via util.Unzip with no limit on the decompressed output size. The only built-in size guard, protocol.MaxMessageLength, is checked against the compressed on-the-wire frame length, not the decompressed size, so it provides no protection. Because decoding (and decompression) occurs in readRequest before authentication, a single unauthenticated connection can send a small (under 2 MB) gzip-compressed message that expands to gigabytes of heap allocation, leading to out-of-memory conditions and service unavailability.

PasswordPusher before 2.8.1 accepts data URI schemes in URL push payloads due to insufficient validation in the valid_url function. Attackers can create malicious pushes containing data:text/html URIs that execute arbitrary JavaScript in victims' browsers when clicked, enabling phishing and credential theft under the trusted PasswordPusher domain.

9Router through version 0.4.41 contains an unauthenticated access vulnerability that allows remote attackers to interact with provider management API endpoints by sending requests without any credentials due to missing authentication middleware in the Next.js API routes under src/app/api/providers/*. Attackers can enumerate, create, modify, or delete provider connections to expose partial credentials, OAuth tokens, and API keys, redirect AI traffic to attacker-controlled servers, or cause complete denial of service by deleting all provider connections.