CVE-20XX-?
agent?

The internet is broken right now.

67 critical CVEs without a fix in the last 48 hours. Developers are crying quietly into their IDEs. 149 more already have a fix that someone still has to apply.

757 CVEs · updated 2026-11-18 22:57 UTC

Reset all
  1. CVE-2026-77295

    MyMangaDB is a self-hosted manga collection manager with automatic metadata, MyAnimeList import, Docker support, and collection statistics. Prior to 2.7.2, the unauthenticated /api/v1/database/export handler in backend/app/api/v1/endpoints/database.py returns a ZIP containing the live SQLite database and the cover-image library. The database includes the users table and hashed_password values, allowing a remote attacker to obtain all application data and conduct offline password attacks without a session or token. This issue is fixed in version 2.7.2.

    Disclosure Library FabianRolfMatthiasNoll MyMangaDB

    8.7 HIGH Fix listed
  2. CVE-2026-73969

    AnsibleForms is a self-hosted web application that turns automation into self-service forms. Prior to 6.2.1, AnsibleForms uses a public hard-coded fallback JWT signing secret from server/config/auth.config.js when ACCESS_TOKEN_SECRET is unset. An unauthenticated attacker can sign a self-contained token with privileged claims because server/src/auth/auth_jwt.js verifies the signature without reloading the user from the database, and server/src/lib/middleware.js trusts the token's roles and options for authorization. A forged token can therefore be accepted as an administrator session, resulting in authentication bypass and full privileged API access. This issue is fixed in version 6.2.1.

    Auth & access Web app ansibleguy76 Ansibleforms

    9.8 CRITICAL Fix listed
  3. CVE-2026-76100

    OwnPay is a self-hosted open-source payment gateway automation platform. In 0.2.0 and earlier, modules/gateways/alipay/AlipayGateway.php fails open when an Alipay callback lacks a signature or the merchant has no configured public key because verify(array $callbackData, array $credentials) falls back to an undefined mode credential that defaults to test mode, while verifyWebhook() returns true unconditionally. An unauthenticated request to POST /webhook/{gateway} can provide a pending order reference and matching amount, pass both verification gates without a valid payment signature, and cause the order and corresponding payment-received ledger entries to be marked completed. No fixed version is available as of this review.

    Crypto Web app own-pay OwnPay

    8.2 HIGH Fix listed
  4. CVE-2026-75506

    OliveTin gives safe and simple access to predefined shell commands from a web interface. From 3000.0.0 until 3000.18.1, OliveTin passes attacker-influenceable output and ot_username values into shellAfterCompleted processing in service/internal/executor/executor.go without applying checkShellArgumentSafety, the shell-safety validation used for primary action execution. The same post-execution template path also expands exitCode without that validation, but the advisory does not establish attacker control of the exit-code value. When an action configures shellAfterCompleted with output or ot_username, shell metacharacters from command output, external data processed by an action, or a trusted-header username can become syntax in the command executed by the stepExecAfter path. This can execute arbitrary commands as the OliveTin process user, including when the primary action uses the safer exec-array mode. This issue is fixed in version 3000.18.1.

    Injection Web app OliveTin

    6.6 MEDIUM Fix listed
  5. CVE-2026-76088

    PGPainless is a simple-to-use OpenPGP API based on Bouncy Castle. Prior to 1.7.8 and from 2.0.0 until 2.0.4, pgpainless-sop/src/main/kotlin/org/pgpainless/sop/InlineDetachImpl.kt InlineDetachImpl and pgpainless-core/src/main/kotlin/org/pgpainless/decryption_verification/MessageInspector.kt MessageInspector manually descend PGPCompressedData packets without incrementing or enforcing MAX_LAYER_DEPTH. A remote unauthenticated attacker can supply an untrusted OpenPGP message containing deeply nested compressed packet structures to a downstream application using inline-detach or message inspection, bypass the intended 16-layer limit, and force continued decompression and packet processing. The resulting resource consumption can block processing threads, exhaust heap memory, and crash the consuming application. This issue is fixed in versions 1.7.8 and 2.0.4.

    Denial of service Pgpainless

    7.5 HIGH Fix listed
  6. CVE-2026-75522

    OpenID4VP is an open-source credential parsing library for SD-JWT VC and mDOC/mDL formats. Prior to 0.8.1, src/trust/ChainBuilder.ts ChainBuilder.tryBuild() terminates chain construction when an attacker-controlled certificate Subject Distinguished Name equals a trusted anchor subject, without verifying cryptographic closure to the anchor public key. A remote unauthenticated attacker can supply a self-signed leaf or intermediate that reuses the trusted Subject Distinguished Name and have credentials signed by the attacker key accepted as issued by the trusted authority. The public src/trust/TrustEvaluator.ts TrustEvaluator.evaluate() path compounds the error by matching a terminus through serialNumber and falling back to anchors[0], which can attribute the forged chain to genuine trusted-authority metadata. This issue is fixed in version 0.8.1.

    Crypto Library Openeudi openid4vp

    7.5 HIGH Fix listed
  7. CVE-2026-76007

    IBM Engineering AI Hub 1.0.0 through 1.4.0 could allow a remote attacker to obtain sensitive information and perform unauthorized actions due to improper validation of a redirect URL.

    Other Web app IBM Engineering AI Hub

    8.1 HIGH Fix listed
  8. CVE-2026-73967

    AnsibleForms is a self-hosted web application that turns automation into self-service forms. Prior to 6.2.1, AnsibleForms stores attacker-influenceable job output that server/src/lib/common.js processes through formatOutput without HTML escaping and client/src/components/AppAnsibleOutput.vue renders through v-html. An authenticated user can submit job input that is echoed into persistent output, causing script execution when another user views the job. Additional designer-controlled HTML and expression values reach v-html in client/src/components/AppForm.vue and client/src/components/BsInputForForm.vue without sanitization. Because Content Security Policy is disabled and the JWT is stored in localStorage, script execution can steal a victim's token and take over the account, including an administrator account. This issue is fixed in version 6.2.1.

    XSS Web app ansibleguy76 Ansibleforms

    8.7 HIGH Fix listed
  9. CVE-2026-73968

    AnsibleForms is a self-hosted web application that turns automation into self-service forms. Prior to 6.2.1, AnsibleForms allows a user with repository management privileges to inject shell syntax through repository uri, branch, name, or host values processed by server/src/models/repo.model.js. The Repo.clone and Repo.addKnownHosts paths interpolate these values into git and ssh-keyscan command strings, and server/src/lib/cmd.js executes the strings through Cmd.executeSilentCommand with shell interpretation enabled. Creating or updating a repository can therefore execute arbitrary commands as the AnsibleForms application process. This issue is fixed in version 6.2.1.

    Injection Web app ansibleguy76 Ansibleforms

    7.2 HIGH Fix listed
  10. CVE-2026-69138

    djangocms-versioning is a general-purpose versioning package for django CMS 4 and later. Prior to 2.6.1, the check_discard and check_edit_redirect Conditions on djangocms_versioning.models.Version omit user_can_change, while VersionAdmin.discard_view and VersionAdmin.edit_redirect_view obtain a version by its primary key and rely on those conditions for object-level authorization. With CMS_PERMISSION set to True, the affected views are gated only by is_staff, allowing any staff user to select content without object-level change permission. With the default CMS_PERMISSION value of False, a staff user must also hold the global Django change_ permission, but that model-level permission does not enforce the intended per-object authorization, so the user can delete a draft version and its content object through discard or create an editable draft from a published version through edit_redirect even when the user cannot change that specific content object. This issue is fixed in version 2.6.1.

    Auth & access Web app django-cms djangocms-versioning

    6.5 MEDIUM Fix listed
  11. CVE-2026-71427

    WPGraphQL Content Blocks is a WordPress plugin that extends WPGraphQL to support querying Gutenberg blocks as data. Prior to 4.8.6, the editorBlocks renderedHtml field in includes/Data/ContentBlocksResolver.php uses resolve_content_blocks() to read a post with get_post() without applying post_password_required(). An unauthenticated caller who knows a supported post identifier can therefore retrieve the complete body of a password-protected published post without a post password, nonce, or valid cookie, exposing paid, member-only, preview, or internal content. This issue is fixed in version 4.8.6.

    Disclosure Web app Wpengine wp-graphql-content-blocks

    5.3 MEDIUM Fix listed
  12. CVE-2026-69131

    django-filer is a file and image management application for Django. Prior to 3.5.0, the delete_files_or_folders action in filer/admin/folderadmin.py checks global Django model-level deletion permission but omits check_folder_edit_permissions and check_files_edit_permissions when FILER_ENABLE_PERMISSIONS is enabled. An authenticated staff user who has a global delete_file, delete_folder, or delete_image permission but is limited to read-only access by FolderPermission can therefore delete files, folders, and descendants from the restricted folder. Deployments using the default disabled per-folder permission system are not affected, and the issue does not create a read-confidentiality impact. This issue is fixed in version 3.5.0.

    Auth & access Web app django-cms django-filer

    4.3 MEDIUM Fix listed
  13. CVE-2026-59964

    uftpd is a UNIX FTP and TFTP server. Prior to 2.16, uftpd's compose_path function in src/common.c can overflow the fixed dir[PATH_MAX] stack buffer when the server runs without chroot and prepends the configured home directory to a near-PATH_MAX attacker-supplied path using memmove. A logged-in FTP client, including an anonymous or guest client where enabled, can trigger the out-of-bounds write through commands such as CWD, RETR, STOR, or LIST, causing memory corruption and a reliable service crash. Remote code execution has not been demonstrated. This issue is fixed in version 2.16.

    Memory safety Server Troglobit Uftpd

    5.9 MEDIUM Fix listed
  14. CVE-2026-59975

    SharpIppNext is a .NET Standard library for building Internet Printing Protocol clients and servers. Prior to 4.1.0, server-side SharpIppNext implementations using SharpIppServer.ReceiveRequestAsync or SharpIppServer.ReceiveRawRequestAsync pass incoming data to IppProtocol.ReadIppRequestAsync in SharpIpp/Protocol/IppProtocol.cs, where ReadDocumentStream is enabled by default and the remaining document body is copied with CopyToAsync into a MemoryStream without a size limit. An unauthenticated remote client can send an arbitrarily large IPP document body, exhausting process memory and causing an out-of-memory failure or service termination. Client-only implementations using SharpIppClient are not affected. This issue is fixed in version 4.1.0.

    Denial of service Library Danielklecha SharpIppNext

    7.5 HIGH Fix listed
  15. CVE-2026-55508

    Wavelog is web-based amateur radio logging software. Prior to 2.5, the callsign parameter accepted by POST /oqrs/search_log in application/controllers/Oqrs.php search_log() reaches application/models/Oqrs_model.php search_log(), where it is concatenated directly into a raw SQL WHERE clause without parameter binding. An unauthenticated remote attacker can use UNION-based SQL injection to read arbitrary application database records, including account data, password hashes, API keys, station credentials, reset tokens, QSO data, and station metadata. Disclosure of password reset tokens can be combined with the forgot_password workflow to take over accounts, including administrator accounts. This issue is fixed in version 2.5.

    Injection Web app Wavelog

    9.4 CRITICAL Fix listed
  16. CVE-2026-55506

    Wavelog is web-based amateur radio logging software. Prior to 2.5, the filename parameter accepted by POST /qsl/addQsoToQsl in application/controllers/Qsl.php addQsoToQsl() is stored without rejecting directory separators or relative path components, and POST /qsl/delete reaches application/models/Qsl_model.php deleteQsl(), which passes the planted qsl_images filename to unlink(). A standard authenticated user can associate the malicious filename with the user's own QSO, satisfy the ownership check, escape the per-user QSL image directory, and delete files writable by the PHP worker. Deleting configuration or operational files can cause denial of service, while deleting installer gate files can reopen the unauthenticated installer and enable application identity takeover through attacker-controlled database configuration. This issue is fixed in version 2.5.

    Disclosure Web app Wavelog

    9.6 CRITICAL Fix listed
  17. CVE-2026-55505

    Wavelog is web-based amateur radio logging software. Prior to 3.2.4, the application/controllers/Lotw.php import() workflow passes the lotw_download_url configuration value to curl without scheme validation, accepts file:// content, and writes the response to a predictable path under ./uploads/ that can be retrieved through an unauthenticated HTTP request. A standard user can trigger the LotW import, although changing lotw_download_url normally requires installer administrator access, and related vulnerabilities can remove that prerequisite. The same missing status-validation pattern affects the LotW lotw_users fetch helper and per-user download helper. This exposes files readable by the www-data user, including environment data, mounted secrets, and application configuration. This issue is fixed in version 3.2.4.

    Disclosure Web app Wavelog

    4.9 MEDIUM Fix listed
  18. CVE-2026-55507

    Wavelog is web-based amateur radio logging software. In 3.2.5 and earlier, the Hams of Note feature stores the administrator-supplied hon_url configuration value through application/controllers/Options.php hon_save() and application/models/Update_model.php update_hams_of_note() fetches that URL without restricting the scheme, host, loopback, link-local, or private network destinations. An administrator can cause the Wavelog server to request arbitrary HTTP or HTTPS destinations from its network position and have parsed response content rendered in the administrator interface, providing a read primitive against otherwise unreachable internal HTTP services. The HTTP status requirement prevents file scheme responses from succeeding, so this issue does not provide local file disclosure. No fixed version is available as of this review.

    Disclosure Web app Wavelog

    6.0 MEDIUM Fix listed
  19. CVE-2026-55169

    Open OnDemand is an open-source high-performance computing portal. Prior to 4.1.6 and 4.2.3, the Open OnDemand dashboard includes noVNC passwords in query parameters generated by apps/dashboard/app/helpers/batch_connect/sessions_helper.rb and apps/dashboard/app/views/batch_connect/sessions/connections/_novnc.html.erb. The server records those query strings in per-user PUN access logs, and insufficiently protected log directories can allow another local user to recover and replay a launch URL. Replaying the credential grants access to the running remote desktop and can result in horizontal or vertical privilege escalation according to the privileges of the desktop owner. This issue is fixed in versions 4.1.6 and 4.2.3.

    Disclosure Web app OSC Ondemand

    8.4 HIGH Fix listed
  20. CVE-2026-55171

    Open OnDemand is an open-source high-performance computing portal. Prior to Open OnDemand 4.1.6 and 4.2.3, and prior to ood-ansible 4.2.1, the portal generator defaults OIDCCryptoPassphrase to SHA1(servername), and the templates/ood_portal.yml.j2 role checks a different variable from the documented passphrase setting. Administrators who set ood_oidc_crypto_passphrase can therefore believe a strong secret is configured while the deployment continues to use an attacker-derivable value. In OIDCSessionType client-cookie deployments, an unauthenticated remote attacker can derive the encryption key and forge a mod_auth_openidc_session cookie for another user, while server-cache deployments permit lower-impact tampering with protected OIDC state and nonce cookies. This issue is fixed in Open OnDemand versions 4.1.6 and 4.2.3 and ood-ansible version 4.2.1.

    Crypto Web app OSC Ondemand

    9.1 CRITICAL Fix listed
  21. CVE-2026-55172

    Open OnDemand is an open-source high-performance computing portal. Prior to 4.1.6 and 4.2.3, apps/dashboard/config/initializers/session_store.rb stores dashboard sessions by default in a predictable /var/tmp/<username> directory, or in another location configured by an administrator, but omits PUN user directory ownership and mode 0700 validation. A local attacker can pre-create another user's directory, read the victim's session identifier and CSRF token after login, or stage a predetermined session identifier. ActiveSupport file-store deserialization of attacker-controlled cache data may also occur in the victim's context, enabling file modification or code execution with the victim's privileges. This issue is fixed in versions 4.1.6 and 4.2.3.

    Other Web app OSC Ondemand

    5.4 MEDIUM Fix listed
  22. CVE-2026-132641

    MISP's EventDelegationsController delegateEvent method accepted a user-supplied sharing_group_id parameter without verifying that the requesting user was a member of the referenced sharing group. When a delegation request was submitted with distribution set to 4 (sharing-group distribution), the raw sharing_group_id value was stored directly into the EventDelegation record. The delegation view subsequently rendered the stored sharing group's name back to the requester. Because no authorization check was performed, any authenticated user with delegation permission could enumerate sharing group identifiers (by guessing sequential IDs or other values) and read the names of sharing groups to which their organisation did not belong, including hidden or restricted groups. Hidden, missing, and malformed identifiers were not distinguished in the response, but the successful case still leaked the group name. The vulnerability required an authenticated session with the perm_delegate capability and knowledge of a valid event ID to delegate. The security impact is limited to disclosure of sharing group names (a low-impact confidentiality breach) and does not grant access to the group's contents or modify any data.

    Disclosure MISP

    5.3 MEDIUM Fix listed
  23. CVE-2026-132642

    MISP contains an access control flaw in its tag management endpoints: events/addTag (single and bulk), events/removeTag, and attributes/removeTag. These endpoints loaded their target event or attribute using an unscoped database query (recursive = -1) that did not apply the calling user's ACL visibility rules. As a result, two distinct security issues existed: (1) Information disclosure: the removeTag endpoint returned different responses depending on whether a given tag was present on a hidden event, allowing an authenticated user to enumerate which tags were applied to events they had no visibility into. (2) Unauthorized modification: users belonging to the host organisation, who are permitted to set local tags on records they can see, could add or remove local tags on events and attributes they could not see, because the canModifyTag ACL check operated on the unscoped object without first verifying the caller's visibility. The fix replaces the unscoped find() calls with ACL-aware fetch methods (fetchSimpleEvent and fetchAttributeSimple) that apply the user's visibility before the object is returned, and adds an explicit empty-object check in removeTag so that a hidden event responds identically to a non-existent one. Affected components include the EventsController (addTag, removeTag), the AttributesController (removeTag), and the ACLComponent (canModifyTag). The vulnerability requires an authenticated user with at least the tagger role (perm_tagger) or host-organisation membership.

    Disclosure Server Misp

    5.3 MEDIUM Fix listed
  24. CVE-2026-132643

    MISP contains an information disclosure vulnerability in the Sighting model when the Sightings_anonymise setting is enabled. The listSightings and sightings/restSearch endpoints accepted a caller-supplied org_id filter parameter and applied it directly to the SQL query without verifying whether the requesting user was authorized to filter by that organisation. Because the anonymization mechanism masks the reporter's organisation in the returned records, a non-site-admin user could bypass the anonymization by specifying another organisation's ID in the filter. The query would then return only that organisation's sightings, effectively naming the reporter and defeating the purpose of the anonymization feature. A negated filter (NOT IN) similarly revealed which sightings belonged to the target organisation by exclusion. The vulnerability requires an authenticated user with sighting read permissions (perm_sighting) who is not a site administrator, and the Sightings_anonymise plugin setting must be enabled. The security impact is the disclosure of which organisation reported a given sighting, undermining the privacy guarantees that the anonymization feature is designed to provide. This issue was reported by ENISA.

    Disclosure Web app Misp

    5.3 MEDIUM Fix listed
  25. CVE-2026-132559

    When payload encryption is enabled, failure messages and stack traces that temporal-proxy is configured to keep from the upstream Temporal service can reach it in plaintext, where anyone who can read the Workflow's history at the upstream can read them. When the proxy returns a failure to a Worker, it decrypts the failure's encoded attributes and copies the message and stack trace back onto the failure, keeping the attributes. A Worker that sends that failure back through the proxy, as SDKs do when an Activity failure becomes the cause of a Workflow failure, sends the restored text with it, and the proxy forwards the text unchanged because the failure already has encoded attributes. The upstream then stores the text in plaintext beside its encrypted copy. This affects deployments that enable failure encryption in the proxy, and deployments that leave it off while the SDK encodes failure attributes itself. Payloads are not affected. temporal-proxy versions 0.9.0 through 1.2.0 are affected. The issue is fixed in 1.0.2, 1.1.2, and 1.2.1, and releases before 1.0.0 receive no fix. Upgrade to the fixed release for the deployed minor release line.

    Other Library Temporal Technologies, Inc. temporal-proxy

    2.3 LOW Fix listed
  26. CVE-2026-132636

    MISP contains an authorization bypass in its event search and filter functionality. When a user performed searches through events/restSearch (value, value prefix, searchall, type, uuid, attribute tags, object filters), objects/restSearch, or the event index tag, galaxy, and quick filters, the underlying SQL subqueries matched against attributes and objects without enforcing the per-attribute distribution, sharing-group, or soft-deleted visibility rules that apply when fetching individual elements. A user who could view an event (for example, a community-distributed event) could therefore use search to determine whether organisation-only, sharing-group-restricted, or soft-deleted attributes and objects existed within that event. Because the value filter supports prefix matching (LIKE), an attacker could enumerate the full value of a hidden attribute one character at a time. Tags attached exclusively to hidden attributes also leaked through the tag, galaxy, and quick-filter paths on the event index. The vulnerability affects the EventsController quick-filter and index-filter logic, the Event model wildcard/value/uuid/tag filter methods, the MispAttribute tag filter, the MispObject buildFilterConditions and restSearch methods, and the ReportFromEvent object filter. Site administrators and users belonging to the event's own organisation retain full visibility, and sync users continue to match soft-deleted elements as intended. The issue was reported by ENISA.

    Disclosure MISP

    5.3 MEDIUM Fix listed
  27. CVE-2026-132637

    MISP contains an authorization bypass in the correlations/eventCorrelations endpoint and the underlying correlation engines (Default and OnDemand). The CorrelationsController::eventCorrelations action did not verify that the requesting user was permitted to view the event specified by the event ID parameter. Any authenticated user could supply the ID of an event belonging to another organisation (distribution level 0, organisation-only) and receive its full correlation results, including attribute identifiers and values. Additionally, for events the user could partially view, the correlation engines filtered only the related (target) side of each correlation row and did not vet the queried event's own source attributes. As a result, correlations keyed by attributes hidden from the user (organisation-only attributes, attributes in sharing groups the user is not a member of, or attributes inside organisation-only objects) were returned with their IDs and values. The Default engine's related-event listing (used in events/view) also included events related solely through such hidden attributes. The vulnerability was reported by ENISA. Successful exploitation requires a valid authenticated session with any role and knowledge or enumeration of a target event ID. The security impact is disclosure of threat-intelligence attribute values (IP addresses, domain names, etc.) and attribute identifiers that the user is not authorised to see, as well as confirmation of the existence of otherwise invisible events.

    Auth & access Misp

    5.3 MEDIUM Fix listed
  28. CVE-2026-132144

    MISP contains an authorization flaw in the user login-email change logic. When an external authentication backend is configured (LDAP, Shibboleth, ApacheSecureAuth, or CustomAuth), the login email address serves as the key by which the directory service resolves a user's identity, organization, and role on each login. The __canChangeLogin() method in UsersController only checked the MISP.disable_user_login_change setting and site-admin status; it did not account for whether an external authentication backend was active. As a result, any authenticated non-admin user (including org admins) could change their own login email to an address that matches a higher-privilege account in the external directory. On the user's next login, the external backend would resolve that email to the target directory identity, granting the attacker the organization and role associated with that identity. This constitutes an account-identity capture and privilege escalation: the attacker effectively assumes the identity of any directory account whose email they can guess or discover. The vulnerability requires an authenticated session with a valid API key or session token and a MISP instance configured with at least one external authentication backend. The impact is limited to the scope of the captured directory identity's privileges within MISP.

    Auth & access Web app Misp

    8.3 HIGH Fix listed
  29. CVE-2026-132140

    A security vulnerability has been detected in DrayTek Vigor2960 up to 1.5.1.6. Affected by this vulnerability is an unknown functionality of the file mainfunction.cgi of the component Apmcfg Upload. The manipulation leads to null pointer dereference. It is possible to initiate the attack remotely. The exploit has been disclosed publicly and may be used. Upgrading to version 1.5.1.7 addresses this issue. Upgrading the affected component is advised. The vendor recommends: "1.5.1.6 isn't the latest version of 2960, the latest version fixes many security issues. And we have another one in our release plan which should solve the issues reported."

    Denial of service DrayTek Vigor2960

    7.7 HIGH Fix listed
  30. CVE-2026-132137

    A vulnerability was identified in Relaticle up to 3.10.0. This affects the function mapToField of the file packages/ImportWizard/src/Livewire/Steps/MappingStep.php of the component Import Wizard. Such manipulation of the argument Target leads to authorization bypass. The attack can be executed remotely. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure.

    Auth & access Web app n/a Relaticle

    2.1 LOW No fix listed
  31. CVE-2026-132139

    A weakness has been identified in DrayTek Vigor2960 up to 1.5.1.6. Affected is the function setSWMProfile of the file /cgi-bin/mainfunction.cgi. Executing a manipulation of the argument filename can lead to stack-based buffer overflow. The attack may be performed from remote. The exploit has been made available to the public and could be used for attacks. Upgrading to version 1.5.1.7 is able to address this issue. It is recommended to upgrade the affected component. The vendor recommends: "1.5.1.6 isn't the latest version of 2960, the latest version fixes many security issues. And we have another one in our release plan which should solve the issues reported."

    Memory safety DrayTek Vigor2960

    8.6 HIGH Fix listed
  32. CVE-2026-130092

    A caller who can reach the temporal-proxy gRPC listener, with or without a valid credential, can grow the proxy's memory until the process is terminated by sending requests with many distinct gRPC method names. temporal-proxy labels its request metrics with the method name exactly as the caller sent it, and records every request, including those that authentication or routing reject. Each distinct method name adds metric series that are never removed. The series grow whether or not the metrics endpoint is scraped, and metric labels configured from request metadata multiply the growth. temporal-proxy versions 0.1.0-alpha.20260716 through 1.2.0 are affected. The issue is fixed in 1.0.2, 1.1.2, and 1.2.1, and releases before 1.0.0 receive no fix. Upgrade to the fixed release for the deployed minor release line.

    Denial of service Temporal Technologies, Inc. temporal-proxy

    8.7 HIGH Fix listed
  33. CVE-2026-130090

    When two templated upstreams in temporal-proxy render the same address, a request routed to one upstream can be sent under the other upstream's credential, namespace translation, and payload encryption keys. temporal-proxy pools connections for templated upstreams by rendered address and TLS server name only, although each upstream's outbound credential, namespace translation, and payload encryption keys, where configured, are attached to its connection. A request routed to one upstream can therefore be sent over the connection the other created, and the upstream Temporal service authorizes it as the other upstream's principal. Configuration validation rejects only identical hostPort values, so templates that differ in text but render the same address are accepted. This mixes the two upstreams' ordinary traffic, and a caller that an external authenticator restricts to one upstream's namespaces can use it to act under the other upstream's credential. Deployments with at most one templated upstream are not affected. temporal-proxy versions 0.1.0 through 1.1.1 are affected. The issue is fixed in 1.0.2, 1.1.2, and 1.2.0, and releases before 1.0.0 receive no fix. Upgrade to the fixed release for the deployed minor release line.

    Auth & access Server Temporal Technologies, Inc. temporal-proxy

    2.3 LOW Fix listed
  34. CVE-2026-130091

    A caller with write permission in a namespace can terminate temporal-proxy through memory exhaustion by storing highly compressed payloads in Workflow history, and the stored history causes the same failure on every later read. temporal-proxy 1.2.0 decompresses every payload marked as compressed in the responses it forwards, whether or not compression is configured. Each payload is limited to 4 MiB after decompression, but the total for a response is not, and the decoded payloads are held in memory together. The caller can store many small, highly compressed payloads through the proxy or directly at the upstream Temporal service. When that history is read through the proxy, for example by a Worker processing a task, the proxy can run out of memory and be terminated. The codec server's /decode endpoint is affected in the same way when it is enabled. Only temporal-proxy 1.2.0 is affected; earlier versions do not decompress payloads. The issue is fixed in 1.2.1. Upgrade to 1.2.1 or later.

    Denial of service Server Temporal Technologies, Inc. temporal-proxy

    7.1 HIGH Fix listed
  35. CVE-2026-130089

    A caller who can send requests to the temporal-proxy gRPC listener can obtain the static API key of an upstream whose address is templated from the request namespace, and then use that key directly against the upstream Temporal service. temporal-proxy does not validate the request namespace before rendering it into a templated upstream address, so when an upstream's hostPort is templated from the namespace, as in the published Temporal Cloud configuration example, a crafted namespace changes the host the proxy dials. If that upstream verifies TLS against the system trust store and sets no literal server name, a caller-controlled host with a publicly trusted certificate passes verification and receives the key. The built-in static-token and JWKS authenticators do not restrict namespaces, so any caller they admit can do this; an external authenticator that admits only listed namespaces prevents it. temporal-proxy versions 0.1.0 through 1.2.0 are affected. The issue is fixed in 1.0.2, 1.1.2, and 1.2.1, and releases before 1.0.0 receive no fix. Upgrade to the fixed release for the deployed minor release line, and rotate any upstream API key that untrusted callers could reach.

    Disclosure Server Temporal Technologies, Inc. temporal-proxy

    7.3 HIGH Fix listed
  36. CVE-2026-127222

    SGLang through 0.5.23, fixed in commit 7b5fb39, contains a missing authorization vulnerability in the POST /load_lora_adapter_from_tensors HTTP endpoint, which lacks the @auth_level(AuthLevel.ADMIN_OPTIONAL) decorator applied to its sibling admin endpoints, allowing unauthenticated network-reachable attackers to invoke it without an API key. Attackers can load a LoRA adapter from attacker-supplied serialized tensors to manipulate the served model, and when combined with a separately documented deserialization bypass, achieve unauthenticated remote code execution as root.

    Auth & access sgl-project Sglang

    8.7 HIGH Fix listed
  37. CVE-2026-128974

    Cleartext storage of the WhatsApp Noise static private key in the shared application container (cck.dat) in WhatsApp Messenger for iOS allows an attacker with access to a full filesystem extraction or equivalent application data copy to recover long-term authentication material and impersonate the victim account.

    Disclosure Device Meta Platforms, Inc. WhatsApp Messenger For iOS

    5.4 MEDIUM No fix listed
  38. CVE-2026-114585

    IBM Engineering AI Hub 1.0.0 through 1.4.0 could allow a remote authenticated attacker to obtain sensitive information due to a server-side request forgery.

    Disclosure Web app IBM Engineering AI Hub

    7.6 HIGH Fix listed
  39. CVE-2026-111516

    Dell Alienware Command Center (AWCC), versions prior to 6.14.105.0, contain a Least Privilege Violation vulnerability. A low privileged attacker with local access could potentially exploit this vulnerability, leading to Elevation of Privilege and Denial of Service.

    Auth & access Dell Alienware Command Center (AWCC)

    6.6 MEDIUM Fix listed
  40. CVE-2026-111517

    Dell Alienware Command Center (AWCC), versions prior to 6.14.105.0, contain a Deserialization of Untrusted Data vulnerability. A low privileged attacker with local access could potentially exploit this vulnerability, leading to Arbitrary Code Execution.

    Code execution Dell Alienware Command Center (AWCC)

    4.4 MEDIUM Fix listed
  41. CVE-2026-111518

    Dell Alienware Command Center (AWCC), versions prior to 6.14.105.0, contain an Improper Restriction of Rendered UI Layers or Frames vulnerability. A high privileged attacker with remote access could potentially exploit this vulnerability, leading to Denial of Service and Launch of Phishing Attacks.

    Denial of service Dell Alienware Command Center (AWCC)

    3.1 LOW Fix listed
  42. CVE-2026-111514

    Dell Alienware Command Center (AWCC), versions prior to 6.14.105.0, contain an Incorrect Default Permissions vulnerability. A low privileged attacker with local access could potentially exploit this vulnerability, leading to Arbitrary Code Execution and Elevation of Privilege.

    Auth & access Dell Alienware Command Center (AWCC)

    7.3 HIGH Fix listed
  43. CVE-2026-111515

    Dell Alienware Command Center (AWCC), versions prior to 6.14.105.0, contain an Improper Authorization in Handler for Custom URL Scheme vulnerability. An unauthenticated attacker with remote access could potentially exploit this vulnerability, leading to Command Execution and Launch of Phishing Attacks.

    Other Dell Alienware Command Center (AWCC)

    7.1 HIGH Fix listed
  44. CVE-2026-111519

    Dell Alienware Command Center (AWCC), versions prior to 6.14.105.0, contain an Improper Authorization in Handler for Custom URL Scheme vulnerability. A high privileged attacker with remote access could potentially exploit this vulnerability, leading to Client-Side Request Forgery and Information Disclosure.

    Disclosure Dell Alienware Command Center (AWCC)

    2.0 LOW Fix listed
  45. CVE-2026-111520

    Dell Alienware Command Center (AWCC), versions prior to 6.14.105.0, contain a Time-of-check Time-of-use (TOCTOU) Race Condition vulnerability. A low privileged attacker with local access could potentially exploit this vulnerability, leading to Code Execution and Elevation of Privilege.

    Other Dell Alienware Command Center (AWCC

    7.3 HIGH Fix listed
  46. CVE-2026-111513

    Dell Alienware Command Center (AWCC), versions prior to 6.14.105.0, contain an Improper Verification of Cryptographic Signature vulnerability. A low privileged attacker with local access could potentially exploit this vulnerability, leading to Arbitrary Code Execution and Elevation of Privilege.

    Crypto Dell Alienware Command Center (AWCC)

    7.3 HIGH Fix listed
  47. CVE-2026-110521

    IBM Engineering AI Hub 1.0.0 through 1.4.0 is vulnerable to cross-site scripting. This vulnerability allows an unauthenticated attacker to embed arbitrary JavaScript code in the Web UI thus altering the intended functionality potentially leading to credentials disclosure within a trusted session.

    XSS Web app IBM Engineering AI Hub

    6.1 MEDIUM Fix listed
  48. CVE-2025-36441

    IBM Financial Transaction Manager (FTM) 3.2.13.0 through 3.2.13.0 Interim Fix 006 stores potentially sensitive information in environment variables that could be read by a local user.

    Disclosure IBM Financial Transaction Manager (FTM)

    6.2 MEDIUM Fix listed
  49. CVE-2026-102434

    IBM Engineering AI Hub 1.0.0 through 1.4.0 could allow a remote attacker to obtain sensitive information due to a Cross Origin Resource Sharing (CORS) misconfiguration.

    Disclosure Web app IBM Engineering AI Hub

    5.3 MEDIUM Fix listed
  50. CVE-2026-76807

    Todoist MCP Server provides tools that allow AI agents to access and modify a Todoist account. Prior to 10.3.3, the optional todoist-mcp-http server accepts browser-origin requests to POST /mcp without validating the Host and Origin headers or authenticating the browser caller. A malicious website can use DNS rebinding to reach the loopback listener and invoke MCP tools through StreamableHTTPServerTransport, which uses the server-held Todoist API key. This exposes Todoist account data and permits operations such as task creation, updates, completion, and deletion. The stdio transport is not browser-reachable and is not affected by this attack path. This issue is fixed in version 10.3.3.

    Auth & access Browser & mail Doist todoist-mcp

    9.3 CRITICAL Fix listed