The internet is broken right now.
47 critical CVEs without a fix in the last 48 hours. Somewhere a staging server is public. It always is. 70 more already have a fix that someone still has to apply.
-
CVE-2026-129301
mail-parser is a tool that parses emails by enhancing the Python standard library and extracting message details into a comprehensive object. From 4.6.0 until 4.6.3, write_attachments() calls _safe_attachment_filename() and write_sample() without isolating failures for individual attacker-controlled MIME parts. An RFC 2231 filename containing a NUL, a basename exceeding NAME_MAX, or an undecodable base64 payload can raise ValueError, OSError, or binascii.Error and abort the entire save loop. Later attachments remain unwritten; when a caller catches only MailParserError, the failure is not surfaced through the expected library-exception path. If a downstream scanner or sandbox reads only the extraction directory, a payload after the crafted part can be missed even though it remains in the delivered message, and the aborted batch also causes availability loss. This issue is fixed in version 4.6.3.
-
CVE-2026-17225
IBM API Connect 10.0.8.0 through 10.0.8.12 and 12.1.0.0 through 12.1.1.2 could allow a remote authenticated attacker to bypass security restrictions and access unauthorized analytics queries due to improper input validation. Request parameters are not validated sufficiently, which could allow restrictions on analytics data to be circumvented.
-
CVE-2026-110970
Płatnik uses HTTP protocol to download updates, libraries and support files. It allows an attacker to conduct a Man-In-The-Middle attack, obtain transmitted data and replace the original files with their own. This issue was fixed in metric 334 in version 10.02.002.
-
CVE-2026-119293
FreeSWITCH is a Software Defined Telecom Stack enabling the digital transformation from proprietary telecom switches to a software implementation that runs on any commodity hardware. In 1.11.2 and earlier, rtp_common_read() in src/switch_rtp.c uses one session-wide autoadj_tally to learn RTP sources that differ from the address negotiated in SDP. An unauthenticated remote attacker who sends plaintext RTP during call setup can make packets from one or more non-advertised sources reach the threshold before the legitimate endpoint, causing FreeSWITCH to adopt the source of the threshold-crossing packet as the media peer. The attacker can then receive call audio, inject media, and deny media to the legitimate endpoint without being on the media path. FreeSWITCH 1.11.2 adds two opt-in controls, but both default to zero and the release remains affected unless both are configured before RTP session creation. This vulnerability is fixed in 1.11.2.
-
CVE-2026-126641
A host key verification flaw was found in ansible-pylibssh and in the ansible.netcommon libssh connection plugin. When trust-on-first-use is enabled (host_key_auto_add on the libssh connection plugin, or AutoAddPolicy in ansible-pylibssh), a host that is already present in known_hosts but now presents a different key is accepted without warning. The pylibssh Session.verify_knownhost() method passes the SSH_KNOWN_HOSTS_CHANGED and SSH_KNOWN_HOSTS_OTHER states to the same missing_host_key() policy hook used for unknown hosts, and the bundled AutoAddPolicy/RejectPolicy classes (as well as ansible.netcommon's MyAddPolicy, which is modeled on paramiko's AutoAddPolicy) therefore add the changed key and continue the connection rather than aborting as paramiko's BadHostKeyException path would. As a result the controller authenticates to the substitute host, runs the play there, and appends the new key to known_hosts, which can disclose connection credentials to an attacker performing a machine-in-the-middle attack. Default configurations (host_key_checking enabled, host_key_auto_add unset) are not affected; non-interactive runs fail closed.
-
CVE-2026-105541
Use of out-of-range pointer offset in Windows Audio Service allows an authorized attacker to elevate privileges locally.
-
CVE-2026-118504
The web interface blocks issue and pull request changes on archived repositories, but several issue routes of the REST API did not check whether the repository is archived. A user allowed to change issues of an archived repository could use the API to edit, close or delete issues, change labels and deadlines, add or delete tracked time and stopwatches, add reactions, change issue blocks, and pin issues, altering data that archiving is meant to freeze. Gitea 28.2.0 added the check to the issue edit, delete and label routes; Gitea 28.3.0 applies it to all modifying repository API requests.
-
CVE-2026-118550
Use of Web Browser Cache Containing Sensitive Information vulnerability in Drupal AI (Artificial Intelligence) allows Exploiting Incorrectly Configured Access Control Security Levels. This issue affects AI (Artificial Intelligence) versions: from 0.0.0 to 1.4.11, from 1.5.0 to 1.5.2.
-
CVE-2026-128637
The MISP Modules website component contains an open redirect vulnerability in its login endpoint. The /login route accepted a user-supplied next query parameter and passed it directly to Flask's redirect() function without validating the destination. An attacker could craft a URL such as https://victim-host/login?next=https://attacker.example/phish and deliver it to a user (e.g., via email or a lookalike page). After the user successfully authenticated, the application issued a 302 redirect to the attacker-controlled external URL. This enables phishing attacks where a user, believing they are still on the legitimate MISP Modules site, may enter additional credentials or sensitive information on the attacker's page. It can also be used to obscure the true destination of a link, bypassing user trust in the legitimate domain. The vulnerability requires the victim to interact with the crafted URL and complete a login, but no prior authentication or special configuration is needed on the attacker's side. The affected component is the website sub-project of misp-modules, specifically the account blueprint in website/app/account/account.py. No version boundaries are specified in the patch metadata.
-
CVE-2026-119398
Time-of-check time-of-use (toctou) race condition in Windows Installer allows an authorized attacker to elevate privileges locally.
-
CVE-2026-123370
compose-lint is a Docker Compose security linter that identifies dangerous misconfigurations before deployment. From 0.9.0 until 0.18.0, compose-lint's YAML parser recognizes lone carriage returns and Unicode line separators U+0085, U+2028, and U+2029 as line breaks while fix position conversion counts only newline characters. A pull request can place one of these separators in a quoted value so fix --apply shifts each later edit and deletes the line after the selected target, including an unselected network or other security setting. The resulting Compose file can remain valid, post-edit checks can pass, and the command can report a successful fix even though the dry-run diff was the only place showing the extra deletion. This issue is fixed in version 0.18.0.
-
CVE-2026-129885
Unrolled Secure before 1.18.0 contains an SSLRedirect bypass vulnerability in Secure.isSSL() that trusts the client-supplied r.URL.Scheme to determine HTTPS transport. Attackers can send plaintext HTTP requests with an absolute-form https:// request-target to skip the HTTPS redirect and have application responses served unencrypted.
-
CVE-2026-122182
Improper link resolution before file access ('link following') in Windows Maps allows an authorized attacker to elevate privileges locally.
-
CVE-2026-104316
IBM Langflow OSS 1.0.0 through 1.12.3 could allow an authenticated attacker to cancel any victim's running background workflow job by sending a single HTTP request, resulting in loss of in-progress compute work with no flow-level permission check.
-
CVE-2026-130743
Grav from 2.0.0 before 2.2.18 contains an insecure blueprint trust vulnerability that allows publisher accounts to read secrets and execute PHP by planting or selecting attacker-controlled blueprint files. Attackers can abuse the import@ directive, uploaded YAML media named as the page template, or form.flash_folder to leak password hashes and keys or invoke static PHP methods for code execution.
-
CVE-2026-63434
MyTube is a self-hosted downloader and player for several video websites. Prior to 1.10.6, when statisticsEnabled and statisticsTrackVisitorActivity are enabled, a visitor-role user can submit events to the statistics ingest endpoint handled by ingestEvents in backend/src/controllers/statisticsController.ts. The handler reads req.body.events and accepts the client-controlled platform, sourceKind, and sessionId fields without server-side derivation or validation before passing the events to the analytics persistence path. A visitor can therefore attribute fabricated events to arbitrary platforms, sources, and sessions, corrupting administrator-visible view counts, download counts, platform breakdowns, source attribution, and session analytics. The server-derived actorRole and surface fields are not forgeable through this issue. This issue is fixed in version 1.10.6.
-
CVE-2026-116589
Datadog dd-trace-py is the Datadog Python APM client. From 3.13.0 until 4.15.5, the IAST string-processing instrumentation does not properly handle certain internal errors while old-style % string formatting processes attacker-controlled input. An unhandled error can terminate the Python process and interrupt the work it is handling. Exploitation requires IAST to be enabled, which is not the default, and application code to apply the affected formatting operation to attacker-controlled input. This issue is fixed in version 4.15.5.
-
CVE-2026-45016
EGroupware is a web-based groupware server written in PHP. Prior to 23.1.20260601 and 26.5.20260507, the processURL2InlineImages logic in api/src/Mail.php treats non-HTTP image sources in an HTML mail body as local paths and passes a decoded attacker-controlled file:// URI to file_get_contents. Any authenticated user with mail access can cause files readable by the web server process to be embedded as inline MIME attachments in outgoing email, exposing local configuration, credentials, private keys, or other server files. This issue is fixed in versions 23.1.20260601 and 26.5.20260507.
-
CVE-2026-122053
The Symfony Serializer component ObjectNormalizer treats any attribute name containing . (and, from Symfony 6.4, [) as a writable property path, allowing untrusted input to modify properties of nested objects and bypass their #[Ignore] metadata attribute in Symfony versions 5.4.41 prior to 5.4.56, 6.4.9 prior to 6.4.50, 7.0.9 prior to 7.1.0, 7.1.2 prior to 7.4.23, and 8.x prior to 8.1.11. ObjectNormalizer::isAllowedAttribute() accepts such names, and setAttributeValue() hands the name straight to PropertyAccessor::setValue(), which parses it as a property path and traverses into the nested object. A payload key such as customer.admin writes to $order->customer->admin. The #[Ignore] metadata attribute is only consulted when the nested object is denormalized in its own right, never when it is reached through a path. #[Groups] restrictions are not bypassed: the path is only accepted when no groups are passed in the context, and then groups restrict nothing. When the root class carries neither #[Ignore] nor #[Groups], no allowlist applies and every key of the payload is let through. Because PropertyAccessor also resolves a path segment against public methods, a segment matching a public method name invokes that method with the attacker’s value.Applications that bind untrusted input to objects are affected, including controllers using #[MapRequestPayload] or #[MapQueryString], whose default normalizer is ObjectNormalizer. With OBJECT_TO_POPULATE, as used for partial updates, the traversal reaches an already hydrated entity graph, and the modified field may be persisted. The fix accepts a property path only when the class metadata declares an attribute of that exact name. Declared paths keep working unchanged, and a path in the payload with no matching declaration is ignored.
-
CVE-2026-103864
Improper output neutralization for logs in Apache Sling Engine. This issue affects Apache Sling Engine: before 3.0.4. Users are recommended to upgrade to version 3.0.4, which fixes the issue.
-
CVE-2026-114727
Roundcube Webmail before 1.6.20 and 1.7.x before 1.7.5 passes potentially unsafe content into ImageMagick when resizing images. SVG and EPS were accepted as valid image types for contact photo upload and conversion, and both formats can reference external resources and invoke ImageMagick delegates when processed. Upstream fixed this in the commit “Fix passing potentally unsafe attachment content into ImageMagick on image resize”, which removes svg and eps from the accepted types.
-
CVE-2026-110941
Mattermost Desktop App versions <=6.3 6.0.3 6.3.2.0 fail to monitor secondary browser windows in the local network access filter which allows an attacker-controlled server to access private network resources via opening pop-out or widget windows. Mattermost Advisory ID: MMSA-2026-00778
-
CVE-2026-127036
Improper neutralization of CRLF sequences ('CRLF injection') vulnerability in ibi WebFOCUS. This issue affects WebFOCUS: 9.1, from 9.2 through 7, and from 9.3 through 7.
-
CVE-2026-120817
A race condition in the schema validation component of MongoDB Server can occur when concurrent writes are checked against a collection validator that uses geospatial query operators. The condition can be caused by an authenticated user with write privileges, and may result in the server terminating unexpectedly.
-
CVE-2026-114726
Roundcube Webmail before 1.6.20 and 1.7.x before 1.7.5 is affected by a mail header injection vulnerability in the Bounce (resend) action. rcube_mime::decode_address_list() did not strip control characters from the decoded display name, so a received message whose From display-name carries CRLF sequences in an encoded word could inject arbitrary headers into the message emitted when the user bounces it. Upstream fixed this in the commit “Fix mail header injection via the From display-name in the Bounce (resend) action”.
-
CVE-2026-129134
MRBS before 1.13.0 contains a password reset poisoning vulnerability that allows unauthenticated attackers to redirect reset links by forging the Host header when $url_base is unset. Attackers can request a reset for a victim so AuthDb::notifyUser() emails a link pointing to their host, capturing the reset key when opened to take over accounts including administrators.
-
CVE-2026-121946
zbateson mail-mime-parser before 3.0.9 and 4.x before 4.0.6 contains a CRLF header injection vulnerability in message-building helpers that fail to strip CR and LF characters. Attackers supplying crafted MIME type, charset, encoding, micalg, or protocol values can inject extra header lines such as a forged Bcc into outgoing messages.
-
CVE-2025-55702
When a client sends a POST with Expect: 100-continue and the origin answers before reading the body, Apache Traffic Server parses the unread body left on the client connection as the next request. An attacker behind a front-end proxy that shares connections to Traffic Server can use this to smuggle requests past the front end and desynchronize responses for other users. This issue affects Apache Traffic Server: from 8.0.0 through 8.1.11, from 9.0.0 through 9.2.15, from 10.0.0 through 10.1.4, 10.2.0. Users are recommended to upgrade to version 9.2.16, 10.1.5, or 10.2.1, which fix the issue.
-
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.
-
CVE-2026-129303
Guard Core is an API security core engine for the FastAPI-Guard, FlaskAPI Guard, and DjAPI Guard adapters. From 1.0.0 until 3.13.0, ContentPreprocessor.preprocess() in guard_core/detection_engine/preprocessor.py collapses all whitespace before built-in pattern matching in the default enhanced mode. This removes carriage returns and newlines required by the http_split and newline-anchored cmd_injection detectors in guard_core/handlers/suspatterns_handler.py, making those detectors unable to match. Framework-decoded CRLF can therefore pass through Guard Core unflagged while the downstream application still receives the raw newline-bearing value. Direct security impact requires a downstream stack or application that persists or re-emits the value into a CRLF-tolerant response or command context. This issue is fixed in version 3.13.0.
-
CVE-2026-126937
Zipline before 4.6.5 contains a TOCTOU race condition in the POST /api/setup endpoint that allows unauthenticated attackers to create multiple SUPERADMIN accounts by racing concurrent requests before the firstSetup flag is reset. Attackers can send simultaneous requests with spoofed X-Forwarded-For headers during the argon2 hashing window to gain full administrative control of the instance.
-
CVE-2026-8639
IBM Security Verify Directory 10.0.0 through 10.0.4 and IBM Verify Directory 11.0.0 is vulnerable to session fixation. An attacker can establish a session and, after the victim authenticates, hijack the authenticated session to gain unauthorized access to the directory management console.
-
CVE-2026-118180
When logging in to a Matrix account, Thunderbird used the homeserver address provided in the server's login response without validating it. A malicious or compromised homeserver could provide a file: URL, causing Thunderbird to send all further Matrix requests of that session to local files or, on Windows, to an attacker-controlled network share, which could leak the user's NTLM credentials. The NTLM credential leak only affects Thunderbird on Windows. Other operating systems are unaffected. This vulnerability was fixed in Thunderbird 159 and Thunderbird 153.6.
-
CVE-2026-115537
A weakness has been identified in gedelumbung HospitalManagement up to c2d45543789a3887067d3915f69d44cfc2cf76a8. This affects the function simpan of the file application/modules/admin/controllers/routing_pages.php of the component URL Validation. This manipulation of the argument url_route causes open redirect. It is possible to initiate the attack remotely. The exploit has been made available to the public and could be used for attacks. This product is using a rolling release to provide continious delivery. Therefore, no version details for affected nor updated releases are available. The project was informed of the problem early through an issue report but has not responded yet.
-
CVE-2026-122149
The PocketTouch and TouchPoint POS protocol includes a `SAVECONFIG` command that appears capable of overwriting till configuration files remotely. An attacker with network access to the POS socket service may be able to submit a crafted `SAVECONFIG` command and replace or modify terminal configuration data.
-
CVE-2026-109873
The User Registration & Membership WordPress plugin before 5.3.2 does not enforce its payment-before-login restriction when a registration is submitted without a membership selection, allowing unauthenticated users to create an account on a payment-gated registration form and log in without completing payment.
-
CVE-2026-118459
A time-derived emergency password can be derived from publicly available application code, enabling authentication without legitimate credentials.
-
CVE-2026-115196
An Open Redirect vulnerability exists in the public form submission endpoint (/form/submit) of Mautic. After a form is submitted, the application redirects the visitor to a return address supplied with the submission without verifying that it points to the Mautic instance or another trusted host. An unauthenticated attacker can exploit this to send visitors from a legitimate Mautic domain to an arbitrary external website, for example to phishing pages that imitate the original landing page or to sites that deliver malware. Any Mautic instance with a published form is affected. Exploitation requires the victim to submit a form prepared by the attacker.
-
CVE-2026-115478
obsidian-tc is a governed, agent-ready Model Context Protocol server for Obsidian vaults. From 1.0.1 until 1.31.4, `create_entity`, `link_entities`, `rename_entity`, `unlink_entities` and `delete_entity` regenerate memory notes from SQLite at the computed `memory/<type>/<name>.md` path without reliably preserving note ownership or the consistency of a failed lifecycle operation. During shared materialization, a hand-authored note or an orphan left by a different entity whose name sanitized to the same path could have its entire body overwritten on the next write naming that type and name, with non-owned frontmatter preserved. `rename_entity` also seeded the destination path with a direct atomic write without validating ownership and could overwrite a foreign note there. Separately, several tools mutated SQLite before fallible filesystem work: `create_entity` and `link_entities` inserted the row or relation before materialization; `rename_entity` updated the entity, including a retired status, before updating its note; `unlink_entities` removed a relation before rematerializing the source note; and `delete_entity`'s cascade removed the target and its relations before rematerializing neighbouring notes. A later filesystem failure could therefore leave a row or relation changed and the memory graph inconsistent with the notes on disk even when the operation failed. An authorized caller with the required memory-tool permissions could, without intending to, destroy the body of a vault note outside the entity being written; `delete_entity` additionally requires its deletion permission and destructive confirmation, which this memory flaw does not bypass. Snapshots-before-write through `restore_note` mitigate the overwrite only where snapshots are enabled and the note had previously been written through obsidian-tc. This issue is fixed in version 1.31.4.
-
CVE-2026-130838
Radicale 3.3.0 before 3.8.3 contains an authorization bypass vulnerability that allows unauthenticated attackers to inherit group-based rights by reusing group memberships never reset between requests. Attackers can send credential-less requests after a group member logs in to read and write group-shared calendar and address book collections evaluated by the from_file rights backend.
-
CVE-2026-128025
FrankenPHP before 1.13.0 contains a path-split interpretation conflict in splitPos in cgi.go that allows unauthenticated attackers to execute protected PHP scripts by embedding .php inside a filename component. Attackers can request paths like /uploads/a.php.txt/b.php to execute uploads/a.php despite route-level or WAF deny rules and forge PATH_INFO, SCRIPT_NAME, and PHP_SELF values.
-
CVE-2026-121425
Nodemailer from 3.0.0 before 10.1.4 contains a parsing flaw in addressparser that ends an RFC 5322 group at a semicolon inside quoted display names or comments. Attackers can submit crafted group addresses that conformant validators approve, redirecting mail to attacker-chosen recipients and bypassing recipient allow lists.
-
CVE-2026-121427
AzuraCast through 0.23.8 fails to remove frontend_config.custom_config in PUT /station/{id}/profile/edit for users lacking the Manage station broadcasting permission, and the stored JSON is written verbatim as key=value lines into the Shoutcast sc_serv.conf file. An authenticated user with only the Manage station profile permission can override adminpassword, password, and portbase, or inject arbitrary directives via newlines in keys, hijacking the station's Shoutcast frontend.
-
CVE-2026-116127
Starlet versions before 0.33 for Perl send a body in responses to HEAD requests, corrupting the next response on a persistent or pipelined connection. The request method is not passed to the response writer, so whatever body the application returns for a HEAD request is sent. With Plack::Middleware::Head the body is empty, but an HTTP/1.1 response without a Content-Length header still gets the chunked encoding terminator. Responses with status 204 or 304 are also sent with their body. Pipelined GET and HEAD requests keep the connection open even when keep-alive is disabled, which is the default. A client or proxy reading the connection takes the extra bytes as the start of the next response, which may desynchronize a proxy that shares the connection across clients.
-
CVE-2026-94380
Weak Password Recovery Mechanism for Forgotten Password vulnerability in HAVELSAN Inc. Liman MYS allows Password Recovery Exploitation. This issue affects Liman MYS: from 2.3.4-1124 before 2.3.5-1131.
-
CVE-2026-126935
Zipline before 4.8.0 contains a race condition vulnerability in src/server/routes/raw/[id].ts that allows remote attackers to bypass file maxViews limits through non-atomic view counting. Attackers holding a file link can send parallel requests that read the same pre-increment count, viewing view-limited or delete-on-max-views files beyond their limit.
-
CVE-2026-107019
Insecure permissions in the feiertage.ImportActivity component of Arbeitszeiterfassung2 versionCode 159 allows unauthorized attackers to create forged holiday/work-time entries via crafted ICS imports.
-
CVE-2026-114648
The MongoDB C# Driver does not neutralize dot and dollar characters in dictionary keys, field names and array-filter identifiers used in LINQ queries before placing them into the field paths it sends to the server. An application user whose input reaches these positions can redirect a query or update to fields the application did not name, which can expose other users' documents and widen updates beyond the intended data. Certain values can also write malformed fields that leave stored documents unreadable by the application until they are repaired.
-
CVE-2026-76406
In Splunk SOAR versions below 8.7.0, an unauthenticated user with access to encrypted credential data from an affected Splunk SOAR instance could submit Basic authentication data to the Representational State Transfer (REST) API to expose and decrypt that encrypted credential data. The vulnerability is possible because Splunk SOAR decrypts encrypted Basic authentication values before completing authentication checks. The attack requires the unauthenticated user to have encrypted credential data from an affected Splunk SOAR instance. The unauthenticated user should not be able to exploit the vulnerability at will. For more information see Using the REST API reference for Splunk SOAR (On-premises) (https://help.splunk.com/en/splunk-soar/soar-on-premises/rest-api-reference/8.5.0/using-the-splunk-soar-rest-api/using-the-rest-api-reference-for-splunk-soar-on-premises) in the Splunk documentation.
-
CVE-2026-84228
IBM Guardium Data Protection 12.2.4 could allow a remote authenticated attacker to obtain sensitive information due to improper input validation.