🚨 CVE-2026-91164
Warpgate is an open source SSH, HTTPS and MySQL bastion host for Linux. From 0.23.0 until 0.27.3, HTTP API token authentication resolves ConfigProvider::validate_api_token into RequestAuthorization::UserToken without enforcing the owning user's allowed_ip_ranges against the trusted client address in warpgate-protocol-http/src/common.rs. An attacker holding a leaked, phished, or exfiltrated X-Warpgate-Token can therefore use it from a prohibited network location. Deployments without allowed_ip_ranges are unaffected, and HTTP target proxying plus SSH, MySQL, PostgreSQL, RDP, VNC, and Kubernetes paths do not accept this vulnerable HTTP token flow. This issue is fixed in version 0.27.3.

🎖@cveNotify

🚨 CVE-2026-91166
Warpgate is an open source SSH, HTTPS and MySQL bastion host for Linux. From 0.25.0 until 0.27.6, the browser SSH path in warpgate-web-ssh/src/manager.rs handles RCEvent::HostKeyUnknown without the presenting hop identity and instead passes ssh_options.host and ssh_options.port for the final target to KnownHosts::trust. In Prompt and AutoAccept modes, a jump host key can therefore be stored for the target address. A machine later presenting the jump host key at the target address can be accepted as the target, allowing interception of user traffic and a newly issued certificate when certificate authentication is used. The native SSH path is unaffected because it tracks each hop separately. This issue is fixed in version 0.27.6.

🎖@cveNotify

🚨 CVE-2026-91167
Warpgate is an open source SSH, HTTPS and MySQL bastion host for Linux. Prior to 0.28.4, PUT /@warpgate/admin/api/users/:id/roles/:role_id reaches api_update_user_role in warpgate-admin/src/api/users.rs through AdminContext but does not require AdminPermission::AccessRolesAssign. A limited administrator with any permission can update expires_at on an existing UserRoleAssignment and clear revoked_at, extending an expiring grant or reinstating a revoked grant. The endpoint cannot create a role assignment that never existed because it returns not found without a matching user and role pair. This issue is fixed in version 0.28.4.

🎖@cveNotify

🚨 CVE-2026-55159
luci-app-adblock-fast a WebUI for fast, lightweight DNS-based ad-blocker for OpenWrt that works with dnsmasq, smartdns, or unbound. Prior to 1.2.4-2, the luci.adblock-fast.setCronEntry RPC method accepts an entry argument containing carriage-return or line-feed characters and serializes it into /etc/crontabs/root as though it were one logical line. An authenticated delegated user with the luci-app-adblock-fast write ACL can therefore create an additional physical root cron entry through applications/luci-app-adblock-fast/root/usr/share/rpcd/ucode/luci.adblock-fast, resulting in persistent command execution as UID 0 when cron runs. The issue is not demonstrated for unauthenticated callers or users without the component write ACL. This vulnerability is fixed in 1.2.4-2.

🎖@cveNotify

🚨 CVE-2026-55897
luci-app-advanced-reboot is a LuCI (web interface) application for OpenWrt that provides a way to reboot your router into an alternative firmware partition or perform reboot operations directly from the web UI. Prior to 1.1.2-6, the luci-app-advanced-reboot read ACL in applications/luci-app-advanced-reboot/root/usr/share/rpcd/acl.d/luci-app-advanced-reboot.json grants rpcd file.exec permission for the general shell interpreter /bin/sh. An authenticated delegated session with that read ACL can supply caller-controlled params; rpcd authorizes the executable path and passes those arguments to the shell, allowing arbitrary commands to execute as root. Builds without the /bin/sh exec grant, including the checked openwrt-24.10 and openwrt-23.05 branches, are not affected by this specific chain. This vulnerability is fixed in 1.1.2-6.

🎖@cveNotify

🚨 CVE-2026-58269
Sync-in Server is an open-source platform for file storage, sharing, collaboration, and syncing. Prior to version 2.4.0, `POST /api/auth/token` authenticates with username and password only, then calls `getTokens()`, which returns full access and refresh JWTs without checking whether the account has TOTP 2FA enabled. An attacker with stolen or phished credentials can bypass 2FA in a single request. The parallel login endpoint (`POST /api/auth/login`) correctly enforces 2FA by calling `setCookies(user, res, true)`, which gates on `user.twoFaEnabled`. Version 2.4.0 patches the issue.

🎖@cveNotify

🚨 CVE-2026-58271
Sync-in Server is an open-source platform for file storage, sharing, collaboration, and syncing. Prior to version 2.4.0, `POST /api/app/sync/register` accepts credentials and a TOTP code to register a desktop sync client. On a failed TOTP, `SyncClientsManager.register()` calls `updateAccesses(user, ip, false)`, which hits a freeze branch that writes `passwordAttempts` back unchanged. The counter never reaches `USER_MAX_PASSWORD_ATTEMPTS` (10) and the lockout gate never fires. A successful guess returns a `{clientId, clientToken}` pair. The token can be exchanged via `POST /api/app/sync/auth/cookie` for a full JWT. While the code is valid, the attacker can also call `POST /api/auth/2fa/disable` to permanently remove MFA. Version 2.4.0 patches the issue.

🎖@cveNotify

🚨 CVE-2026-62247
Supabase Realtime provides Broadcast, Presence, and Postgres Changes via WebSockets. Prior to 2.111.2, Realtime authorization does not correctly honor the per-extension presence.read row-level security policy when a private-channel client is allowed presence.write but explicitly denied presence.read. Under that differential policy, the client can receive presence_diff messages containing other members' presence metadata, including application-defined location, online-status, roster, viewing, or typing information. Deployments with uniform presence visibility have no differential, and postgres_changes row data is unaffected. This issue is fixed in version 2.111.2.

🎖@cveNotify

🚨 CVE-2026-58270
Sync-in Server is an open-source platform for file storage, sharing, collaboration, and syncing. Prior to version 2.4.0, the sync diff endpoint compiles a user-supplied string into a `RegExp` with no complexity validation. A catastrophic-backtracking pattern (e.g. `^(a+)+b`) blocks the Node.js event loop, making the entire server unresponsive to all users until the container is restarted. Version 2.4.0 patches the issue.

🎖@cveNotify

🚨 CVE-2026-61647
NotebookLM MCP is an MCP server and HTTP service for interacting with Google NotebookLM and exporting generated content to local vault directories. Versions 1.6.0 through 2.0.2 contain a path traversal vulnerability in the `POST /batch-to-vault` endpoint, also exposed through the `batch_to_vault` MCP tool beginning in version 1.7.0, because attacker-controlled `vault_dir` and `slug_prefix` values can cause Markdown and JSON files to be written outside the intended vault directory to any location writable by the server process. Version 2.0.3 sanitizes `slug_prefix` and supports vault containment when `NOTEBOOKLM_VAULT_ROOT` is configured; containment is not enabled if that variable is unset. Users unable to upgrade should run the server as a dedicated unprivileged account restricted to the intended vault, keep the HTTP endpoint limited to localhost, and validate `vault_dir` values supplied by LLMs processing untrusted content.

🎖@cveNotify

🚨 CVE-2026-61652
Zapros, a Python HTTP client, prior to version 0.14.0 is vulnerable to denial of service via memory exhaustion. The issue affects all callers who streamed compressed responses relying on the chunk size — explicit (`iter_bytes(chunk_size=...)`) or the default — to bound memory. The decoder ignored that bound, so a chunk could be far larger than requested and a single compressed response could overflow memory. Version 0.14.0 contains a patch. Some workarounds are available. Read the still-compressed body with `Response.iter_raw()` / `Response.async_iter_raw()`, which bypass the built-in decoders, and decompress it yourself with an explicit output-size bound (e.g. `zlib`'s `max_length`), aborting once a configured limit is exceeded. Where feasible, send `Accept-Encoding: identity` to disable response compression so bodies are not decompressed client-side. Avoid decoding response bodies from untrusted servers.

🎖@cveNotify

🚨 CVE-2026-61743
Chartbrew is an open-source web application that can connect directly to databases and APIs and use the data to create charts. Prior to 5.2.2, Chartbrew's server/modules/safeRequest.js calls validateOutboundUrl() to resolve and validate a target hostname, but request-promise performs a separate DNS resolution for the actual connection. An authenticated user who can create or test API connections and controls the target DNS name can return a public address during validation and a private address during the request, bypassing the protections implemented by server/modules/outboundTargetPolicy.js. This DNS rebinding condition can expose internal services, localhost resources, or cloud metadata endpoints. This issue is fixed in version 5.2.2.

🎖@cveNotify

🚨 CVE-2026-61851
Chartbrew is an open-source web application that can connect directly to databases and APIs and use the data to create charts. Prior to 5.2.2, Chartbrew's runQuery() implementation in server/modules/ai/orchestrator/tools/runQuery.js attempts to enforce read-only database access with a blocklist containing only seven SQL keywords. An authenticated user with AI feature access can submit dangerous statements or database functions that are absent from the read-only keyword blocklist, causing them to execute without SQL injection or keyword-obfuscation techniques. Depending on the database engine, configuration, and database-user privileges, this can expose or write files, access internal network resources, change database privileges, execute commands, or alter data. This issue is fixed in version 5.2.2.

🎖@cveNotify

🚨 CVE-2026-61852
Chartbrew is an open-source web application that can connect directly to databases and APIs and use the data to create charts. Prior to 5.2.2, Chartbrew's runQuery() implementation in server/modules/ai/orchestrator/tools/runQuery.js interpolates the AI tool's row_limit parameter into a SQL LIMIT clause without runtime integer validation. The read-only keyword check runs before this value is appended, so an authenticated user who can influence a model-generated non-integer row_limit can add SQL that bypasses the earlier check. Successful exploitation can execute arbitrary statements against the connected database, including reading or changing data and, where database permissions permit, accessing files or executing operating-system commands. This issue is fixed in version 5.2.2.

🎖@cveNotify

🚨 CVE-2026-65980
Chartbrew is an open-source web application that can connect directly to databases and APIs and use the data to create charts. Prior to 5.2.3, Chartbrew's ClickHouse protocol in server/sources/plugins/clickhouse/clickhouse.protocol.js calls applySqlVariables() from server/sources/shared/sql/sql.variables.js without enabling the escapeBackslash option. For a ClickHouse-backed chart with variable binding, an attacker can supply a backslash before a quote so quote doubling does not keep the value within its intended SQL string literal. Public dashboards can expose this path without authentication, and successful exploitation can execute arbitrary ClickHouse SQL to disclose data or, when the database configuration permits, access files or internal network resources. This issue is fixed in version 5.2.3.

🎖@cveNotify

🚨 CVE-2026-75510
Novu provides an API for sending notifications through multiple channels. Prior to 3.18.0, Novu's @novu/js In-App Inbox and the @novu/react Inbox component accept a notification call-to-action redirect.url from the v1 cta.data object and pass it through apps/api/src/app/inbox/utils/notification-mapper.ts and packages/js/src/ui/components/Notification/DefaultNotification.tsx to the navigate function in packages/js/src/ui/context/InboxContext.tsx without validating its URL scheme. An authenticated organization member or environment API-key holder can store a javascript: redirect with target _self in an in-app workflow. When a recipient using a Chromium-based browser clicks the notification, window.open executes the redirect in the current inbox-hosting origin, which can expose session material and permit authenticated actions in a customer application or the self-hosted Novu dashboard. This issue is fixed in version 3.18.0.

🎖@cveNotify

🚨 CVE-2026-75511
Novu provides an API for sending notifications through multiple channels. Prior to 3.18.0, Novu accepts chat webhook URLs from subscriber credentials.webhookUrl, channel endpoint endpoint.url, event payload.webhookUrl, and event overrides.webhookUrl, then passes the selected endpoint.url through send-message-chat.usecase.ts to raw HTTP requests in the Slack, Discord, Mattermost, Microsoft Teams, Grafana On-Call, Ryver, Rocket.Chat, GetStream, and Zulip providers. An authenticated user can supply an internal or otherwise restricted destination because these paths do not apply normalizeOutboundHttpUrl, assertSafeOutboundUrl, or the DNS-pinned safeOutboundJsonRequest protection used by the generic webhook providers. The Novu worker can consequently issue attacker-directed POST requests to internal network services and cause interactions or actions supported by those services. This issue is fixed in version 3.18.0.

🎖@cveNotify

🚨 CVE-2026-75517
Novu provides an API for sending notifications through multiple channels. Prior to 3.18.0, Novu integration mutation use cases including remove-integration, update-integration, auto-configure-integration, and set-integration-as-primary look up an integration by integrationId and organizationId without consistently enforcing environmentId. A caller with access to one environment in an organization can target an integration identifier from another environment and delete the integration, modify its credentials, change the primary provider, or trigger auto-configuration. The repository advisory defines both environment API keys and dashboard sessions as affected. The 3.18.0 change enforces the environment boundary only for environment-scoped API-key authentication, while its tests intentionally retain cross-environment dashboard-session behavior, so it does not completely remediate the advisory-defined scope. Version 3.18.0 is a partial fix attempt, and the dashboard-session behavior requires curator resolution.

🎖@cveNotify

🚨 CVE-2026-77619
Vector is a high-performance observability data pipeline. From 0.15.0 until 0.57.0, the logstash source reads a 32-bit compressed-frame length from the network and uses it to size an in-memory buffer without an upper bound. An unauthenticated remote peer that can reach the default 0.0.0.0:5044 listener can send a minimal frame declaring a multi-gigabyte payload, causing an excessive allocation that can abort Vector or invoke the host OOM killer. Because the allocation follows the declared length rather than bytes transmitted, the attacker has low resource cost, and process termination can halt log ingestion for every tenant on a shared pipeline. This issue is fixed in version 0.57.0.

🎖@cveNotify

🚨 CVE-2026-77620
Vector is a high-performance observability data pipeline. From 0.15.0 until 0.57.0, the logstash source feeds each decompressed frame back into its decoder without limiting nested compression depth. An unauthenticated remote peer that can reach the default 0.0.0.0:5044 listener can send many nested compressed frames, causing recursive decoding that exhausts the worker thread stack and aborts the process. The same nested construction amplifies decompressed input, and process termination can halt log ingestion for every tenant on a shared pipeline. This issue is fixed in version 0.57.0.

🎖@cveNotify

🚨 CVE-2026-77621
Vector is a high-performance observability data pipeline. From 0.10.0 until 0.57.0, the file sink renders its templated path from event fields and opens the result without confining it to an intended directory. When an untrusted source supplies an event field used by the path template, the value can contain an absolute path or parent-directory traversal, causing Vector to create parent directories and create or overwrite files outside the intended location with the Vector process privileges. The resulting file write can modify sensitive files and can lead to code execution when a scheduled task, authorization file, or subsequently executed script is targeted. This issue is fixed in version 0.57.0.

🎖@cveNotify