The internet is broken right now.
38 critical CVEs without a fix in the last 48 hours. Somewhere a staging server is public. It always is. 47 more already have a fix that someone still has to apply.
-
CVE-2026-120342
A vulnerability was detected in mathurvishal CloudClassroom-PHP-Project. The affected element is an unknown function of the file viewquery.php. The manipulation of the argument eid results in sql injection. The attack can be executed remotely. The exploit is now public and may be used. This product implements a rolling release for ongoing delivery, which means version information for affected or updated releases is unavailable. The vendor was contacted early about this disclosure but did not respond in any way.
-
CVE-2026-105894
The Ak Framework plugin for WordPress, included in the Newsy - Viral News & Magazine WordPress Theme, is vulnerable to Privilege Escalation in all versions up to, and including, 2.9.8. This is due to missing capability checks in the ajax_refresh_nonce and ak_admin_ajax handlers before passing attacker-controlled option names and values to update_option(). This makes it possible for unauthenticated attackers to update arbitrary WordPress options including users_can_register and default_role to escalate privileges to administrator by enabling user registration with administrator as the default role.
-
CVE-2026-118986
CallBackery versions before 0.62.0 for Perl allow session cookie forgery via a signing secret generated with rand(). The secret that signs session cookies is built from two calls to the built-in rand(), which is seeded with 32 bits. The secret therefore has at most 2**32 possible values, and one signed cookie is enough to test each candidate offline. An attacker who obtains a single session cookie can recover the secret and sign cookies for any user, including an administrator.
-
CVE-2026-118995
CallBackery versions before 0.62.0 for Perl accept unintended values in form fields checked by createRxValidator. createRxValidator() is documented to anchor the supplied pattern at both ends, but builds /^${rx}$/ without grouping the pattern. An alternation such as yes|no then accepts any value starting with yes or ending with no, and $ also accepts a value with one trailing newline. An authenticated user can submit values the pattern was meant to reject, which an application may then write into generated configuration files.
-
CVE-2026-118664
decode-uri-component provides a more robust alternative to decodeURIComponent. Prior to 0.5.1, decodeUriComponent in index.js falls back to customDecodeURIComponent after native decoding throws, and the fallback collects every distinct percent-encoded run before performing a full-string regular-expression replacement for each entry. Malformed input containing many distinct runs can therefore make both the number of replacements and the scanned input length grow together, causing quadratic CPU consumption and blocking the Node.js event loop; request-body inputs can reach the demonstrated size, while ordinary URL header limits reduce the practical delay. This issue is fixed in version 0.5.1.
-
CVE-2026-81509
nforwardauth is a forward auth service that allows the use of a single authentication middleware for multiple sites. Versions prior to 1.5.1 accept an unvalidated `r` query parameter on the `/login` endpoint and use them as the post-authentication redirect destination. An unauthenticated attacker can create a legitimate `nforwardauth` login URL that redirects a victim to an attacker-controlled website after the victim successfully authenticates. This issue may facilitate phishing and other social-engineering attacks by abusing the trust associated with the legitimate authentication domain. This vulnerability does not provide an unauthenticated ForwardAuth bypass. Version 1.5.1 contains a patch.
-
CVE-2026-116176
vLLM from 0.9.2 before 0.24.0 contains a server-side request forgery vulnerability that allows API callers to force outbound connections by embedding a decode address in the X-Request-Id header. Attackers can make a P2pNcclConnector prefill worker ZMQ-connect to an attacker host, leaking the NCCL unique_id and KV-cache tensors or blocking the worker.
-
CVE-2026-105481
A flaw in Node.js HTTP Agent connection reuse can cause a client to accept as valid a response that was sent before the client has sent the request. This is an incomplete fix of CVE-2026-48931. The guard that destroys idle keep-alive sockets when unexpected data arrives only runs when the event loop polls for I/O. If an unsolicited response has already reached the socket but has not been read yet when the socket is handed to a new request, for example because the event loop was busy with synchronous work, the guard is removed without seeing it and the data is parsed as the response to the new request. This vulnerability affects Node.js **26.x**, **24.x**, and **22.x**.
-
CVE-2026-115730
i18next-http-backend 4.0.2 contains an incomplete fix in interpolateUrl() that allows attackers to redirect translation requests by supplying an ns value starting with a slash. Attackers controlling the namespace value can form a protocol-relative URL in loadPath or addPath templates to load attacker-hosted translations, enabling UI content injection and cross-site scripting.
-
CVE-2026-98911
In the Linux kernel, the following vulnerability has been resolved: dpaa2-switch: fix the error path in dpaa2_switch_rx() In case of an error in dpaa2_switch_rx(), the dpaa2_switch_free_fd() function is called in order to free the FD. This is incorrect since the dpaa2_switch_free_fd() is intended to be used on Tx frame descriptors, meaning that it expects in the software annotation area of the FD data to find a valid skb pointer on which to call dev_kfree_skb(). Fix this by extracting the dma_unmap_page() from dpaa2_switch_build_linear_skb() directly into the dpaa2_switch_rx() function. This allows us to directly use free_pages() in case of an error before an SKB was created and kfree_skb() afterwards.
-
CVE-2025-71434
Weaver E-cology 8.0, 9.0, and 10.0 prior to 10.76 contains an unauthenticated SQL injection vulnerability in the /api/doc/out/more/list endpoint via the elementmore parameter with srcType=2. Attackers can inject SQL through doc_datefrom/doc_dateto fields and read results via /api/ec/dev/table/counts to extract database information. Exploitation evidence was first observed by the Shadowserver Foundation on 2026-07-19.
-
CVE-2026-133074
A flaw was found in Moodle. Insufficient authorization checks in the quiz overrides web service allow an authenticated user with teacher privileges to modify user and group overrides for quizzes they do not have permission to manage. An attacker could exploit this vulnerability to improperly alter exam access, deadlines, or testing conditions across unauthorized courses.
-
CVE-2026-133075
A flaw was found in Moodle. Due to insufficient authorization checks in the quiz random question filter web service, an authenticated user with teacher privileges can modify random question settings in quizzes they do not have permission to manage. This flaw allows unauthorized users to alter assessment configurations across courses they do not oversee.
-
CVE-2026-133098
In JetBrains TeamCity before 2026.2.2 stored XSS in GitHub issue tracker suggestions was possible
-
CVE-2026-133031
In JetBrains TeamCity before 2026.2.1 stored XSS in Git VCS root settings was possible
-
CVE-2026-133033
In JetBrains TeamCity before 2026.2.1 stored XSS in Maven Build Info module links was possible
-
CVE-2026-133035
In JetBrains TeamCity before 2026.2.1 stored XSS on the change Files tab was possible
-
CVE-2026-133036
In JetBrains TeamCity before 2026.2.1 stored XSS in the Issue Trackers health report was possible
-
CVE-2026-129013
Missing Authorization vulnerability in Deepen Bajracharya Video Conferencing with Zoom video-conferencing-with-zoom-api allows Exploiting Incorrectly Configured Access Control Security Levels.This issue affects Video Conferencing with Zoom: from n/a through 4.6.12.
-
CVE-2026-88711
Improper neutralization of special elements used in an OS command in certain ASUS Router firmware models allows an remote attacker to to execute arbitrary commands with root privileges by inducing a logged-in administrator to visit a malicious website. Refer to the ' Security Update for ASUS Router Firmware ' section on the ASUS Security Advisory for more information.
-
CVE-2026-130803
The firmware and flash memory store critical cryptographic secrets, such as global AES keys and network master keys, in unencrypted formats. Additionally, active session tokens (Store Pairing Tokens) are written to the flash in plaintext, exposing live network credentials.
-
CVE-2026-68905
Hugo Blog is a tech blog site. In commit 744ca15 and prior, the AI assistant proxy exposes three HTTP endpoints (POST /api/blog-assistant, POST /api/blog-search-embedding, POST /api/blog-search-rerank) without any authentication check. When deployed following the documented Nginx reverse-proxy setup, any unauthenticated remote attacker can send arbitrary requests that the proxy forwards to the upstream LLM/embedding/rerank API using the operator's server-side API key (OPENAI_API_KEY / AI_SEARCH_API_KEY). This enables unlimited quota consumption, cost exhaustion, and potential service unavailability for the operator. At time of publication, there are no publicly available patches.
-
CVE-2026-127093
Rustrak before 0.26.0 contains a server-side request forgery vulnerability that allows users with integrations update permission to send requests to arbitrary destinations by supplying routing_override.url to POST /api/integrations/{id}/test. Attackers can target loopback, private-network and cloud metadata addresses with custom headers and redirects followed, reading non-2xx response bodies returned in the API response.
-
CVE-2026-126426
OpenEMR 5.0.2 before 8.5.0 contains a missing authentication vulnerability in mfa_registrations.php and mfa_u2f.php that allows session hijackers to delete MFA factors or register U2F keys without password re-entry. Attackers holding a victim's authenticated session can remove all second factors or add their own U2F key, bypassing MFA at later logins with the password.
-
CVE-2026-126770
OpsTree REMS through 1.0.0 contains a server-side request forgery vulnerability that allows authenticated users to make the backend connect to arbitrary destinations through the database Test Connection feature. Attackers can set a database entry's host and port to localhost, internal addresses, or attacker-controlled servers to probe internal hosts and ports reachable from the REMS server.
-
CVE-2026-123267
CAI Content Credentials is affected by an Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal') vulnerability that could result in a Security feature bypass. An attacker could leverage this vulnerability to access unauthorized files or directories outside the intended restrictions. Exploitation of this issue requires user interaction in that a victim must visit a maliciously crafted URL or interact with a compromised web page.
-
CVE-2026-92544
SQL injection in the LLM usage counters of Kong Gateway Enterprise and Kong AI Gateway may allow a client to execute arbitrary SQL against the Kong database via a crafted model name.
-
CVE-2026-119852
Missing authorization in Windows Installer allows an authorized attacker to elevate privileges locally.
-
CVE-2026-60046
An unauthenticated remote attacker can perform password resets for any user with known username and email without rate limiting. This can result in a total loss of availabilit since the password will be reseted.
-
CVE-2026-123475
Missing Authorization vulnerability in PT Norther Lights Production MapSVG mapsvg-lite-interactive-vector-maps allows Exploiting Incorrectly Configured Access Control Security Levels.This issue affects MapSVG: from n/a through 8.16.3.
-
CVE-2026-121760
Missing Authorization vulnerability in weDevs weDocs wedocs allows Exploiting Incorrectly Configured Access Control Security Levels.This issue affects weDocs: from n/a through 2.6.1.
-
CVE-2026-124432
phpMyFAQ 4.1.9 before 4.1.11 and 4.2.0 pre-releases before 4.2.0-beta.3 contain an account enumeration vulnerability in the unauthenticated POST /api/webauthn/prepare-login endpoint, which returns stable stored credential IDs for passkey-registered accounts but an md5(time())-derived fallback ID otherwise. Unauthenticated attackers can compare allowCredentials IDs from requests sent more than one second apart, without rate limiting, to identify existing passkey-registered accounts and their credential IDs.
-
CVE-2026-17372
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 probe internal network resources due to insufficient validation of webhook destination URLs. User-configured webhook addresses are not sufficiently restricted, so the server could be made to contact internal systems.
-
CVE-2026-55649
GeoNode is a geospatial content management system, a platform for the management and publication of geospatial data. Prior to 5.0.3, document abstract metadata is stored without sanitization, allowing a low-privileged remote user to persist arbitrary HTML and JavaScript that executes in another user's browser when the document page is visited. This issue is fixed in version 5.0.3.
-
CVE-2026-9797
A flaw was found in Keycloak's administrative services (org.keycloak.services.resources.admin). A highly privileged attacker, authenticated with a 'manage-clients' role, could exploit this vulnerability through several methods, including manipulating client JSON Web Key Sets (JWKS), swapping client authenticator types and resetting secrets, or downloading legitimate signing keys. Successful exploitation allows the attacker to obtain a token with 'realm-admin' role assignments, leading to privilege escalation and potential unauthorized access to sensitive resources.
-
CVE-2026-121174
Missing authorization in BackgroundFetch in Google Chrome prior to 157.0.8103.38 allowed a remote attacker to bypass web origin policy via a crafted HTML page. (Chromium security severity: High)
-
CVE-2026-88968
Claude Code is an agentic coding tool. Prior to 2.1.251, the Edit tool validates a target using pathname and symlink resolution that is not bound to the underlying directory inode or file descriptor. On macOS, attacker-controlled code running in the Bash sandbox can atomically replace an in-workspace parent directory with a symlink after validation, causing the unsandboxed Edit implementation to overwrite a user-writable file outside the workspace. Exploitation requires attacker-controlled code to run inside the sandbox, such as from a malicious repository or dependency script. This issue is fixed in version 2.1.251.
-
CVE-2026-45146
Chamilo LMS is an open-source learning management system. Prior to 1.11.40 and 2.0.1, an authenticated administrator can call the get_sessions action in /main/inc/ajax/model.ajax.php with _search=true and a crafted JSON filters parameter whose groupOp field contains SQL syntax. Because groupOp is not restricted to the expected AND or OR operators before it is incorporated into query logic, attacker-controlled input is interpreted by the database and enables blind or time-based SQL injection. Successful exploitation can disclose sensitive user, course, session, configuration, password-reset, and administrative data, modify or delete records subject to database privileges, and cause denial of service through expensive delayed queries. This issue is fixed in versions 1.11.40 and 2.0.1.
-
CVE-2026-121401
FreeRDP before 3.32.2 contains a denial of service vulnerability where rdg_process_packet dispatches a PKT_TYPE_EXTENDED_AUTH_MSG without verifying negotiated mode or auth context. A malicious or compromised RD Gateway can send the extended-auth packet before tunnel authentication, triggering a null-pointer dereference that crashes the client.
-
CVE-2026-121396
FreeRDP clients from 3.0.0 before 3.32.2 built with verbose WinPR assertions contain a reachable assertion in rts_read_auth_verifier triggered by malformed RTS padding. Attackers controlling an RD Gateway can send a crafted RPC-over-HTTP RTS PDU with inconsistent authentication padding to abort the connecting client process.
-
CVE-2026-116862
Unauthenticated Broken Access Control in WP REST Cache <= 2026.3.1 versions.
-
CVE-2026-116401
Unauthenticated Broken Access Control in HUSKY <= 1.4.6 versions.
-
CVE-2026-119764
Mythic 3.2.16 through 3.4.0.61 contains a stored cross-site scripting vulnerability that allows attackers controlling target endpoint files to run JavaScript in the Mythic UI origin by planting HTML in .pdf-named files. When an operator downloads the file and clicks Preview Media, content-sniffed HTML from /direct/view renders same-origin, stealing localStorage JWTs for full GraphQL and REST access.
-
CVE-2026-115599
Missing Authorization vulnerability in ThimPress LearnPress learnpress allows Exploiting Incorrectly Configured Access Control Security Levels.This issue affects LearnPress: from n/a through 4.4.11.
-
CVE-2026-114193
Missing Authorization vulnerability in Eyal Fitoussi GEO my WordPress geo-my-wp allows Exploiting Incorrectly Configured Access Control Security Levels.This issue affects GEO my WordPress: from n/a through 4.5.5.5.
-
CVE-2026-116393
Missing Authorization vulnerability in iHomefinder, Inc. Optima Express optima-express allows Exploiting Incorrectly Configured Access Control Security Levels.This issue affects Optima Express: from n/a through 9.0.0.
-
CVE-2026-113487
The GeoDirectory WordPress plugin before 2.8.196 does not properly sanitise or escape a post badge click-through link value before outputting it in an anchor element on a listing, allowing users with the Contributor role and above to perform Stored Cross-Site Scripting attacks against higher-privileged users who view the listing, including an administrator reviewing moderated content.
-
CVE-2026-113488
The GeoDirectory WordPress plugin before 2.8.196 does not properly escape a post badge shortcode attribute for a JavaScript context before placing it in an inline event handler, allowing users with the Contributor role and above to perform Stored Cross-Site Scripting attacks against anyone who clicks the badge, including administrators.
-
CVE-2026-111967
The MP3 Audio Player WordPress plugin before 5.16 does not escape one of its audio player shortcode attributes before outputting it into an HTML attribute, allowing users with Contributor-level access and above to inject arbitrary JavaScript that executes when another user, such as an administrator previewing a pending post, views the content.
-
CVE-2026-107098
The Royal Addons for Elementor WordPress plugin before 1.7.1070 does not sanitise and escape one of its map widget's size settings before concatenating it into an HTML attribute, allowing users with the Contributor role or above to inject JavaScript that runs in the browser of any user who views the page, including an administrator reviewing a pending submission.