Google's Chrome 156 stable release (156.0.8078.47, .48 on some Windows/Mac) fixed 353 security issues, including six critical bugs in Skia, Accessibility, CSS, Paint and WebRTC. On Oct 27 a follow-up update (156.0.8078.73/.74) fixed 52 more, four of them critical (FileSystem, FileAPI, Skia, Layout). On Oct 29 a third update (156.0.8078.79/.80) fixed 21 more, including a critical memory corruption bug in MediaStream (CVE-2026-76024). Google has not said any are exploited. The rollout is staged, so update and relaunch rather than waiting.
2026-11-04: A public GitHub repository publishes a PoC and checker for CVE-2026-118306 (FileSystem use-after-free fixed in Chrome 156.0.8078.73); Google has not reported exploitation. (source)
2026-10-29: Chrome 156.0.8078.79/.80 fixes 21 more security bugs, one critical (CVE-2026-76024, memory corruption in MediaStream) and 19 high; none reported exploited. (source)
2026-10-28: Anthropic says an unreleased Claude model found 62 of the Chrome flaws fixed as of Chrome 156 (all critical or high, 13 sandbox escapes) and chained four into a one-click sandbox escape; lab-built, not seen in attacks. (source)
CVE-2026-121614 affects Docker Sandboxes 0.31.0 before 0.50.0. Host-side Git commands did not neutralize repository-controlled Git configuration. An attacker who can modify a host repository's Git config—for example through a direct-mode sandbox with the repo mounted read-write—can set core.fsmonitor or another setting that invokes an external program. A later clone-mode fetch, or from 0.39.0 an sbx kit push, can run that program on the host with the user's privileges, outside the microVM. Clone-mode's read-only host mount alone cannot poison the config. Fixed in Docker Sandboxes 0.50.0 (released Oct 26 2026; CVE published ~Nov 3). Vendor CVSS 4.0 is 7.1 (High). Upgrade to 0.50.0+; avoid read-write mounts of untrusted repos until patched.
2026-11-03: CVE-2026-121614 is published describing the Git-config sandbox escape in 0.31.0 before 0.50.0. (source)
2026-10-26: Docker Sandboxes v0.50.0 ships: host Git commands no longer run programs configured by a sandbox through the workspace repository's .git directory. (source)
FrankenPHP 1.13.5, released November 5, is a security release fixing 17 vulnerabilities, most of them found by Anthropic's OSS Scanner. The most severe (High, GHSA-j99q-3wr8-xj43) let php_server run an uploaded non-PHP file as PHP, for instance /uploads/photos.php.php/avatar.jpg, so an attacker able to upload a file under the document root could execute code. Several others affect the default php_server configuration: an open redirect, HTTP/2 Host spoofing, a crash a client can trigger with the official Docker images' PHP settings, and slow request bodies that block every PHP thread. Others include Caddy path-rule bypasses with dot segments and encoded slashes, source disclosure with worker match rules, and session output sent to the wrong client in worker mode. The release bundles Caddy 2.11.9, released the same day, which fixes FastCGI issues of its own: clients overriding HTTP_* variables set with env, encoded-slash and Windows 8.3 short-name path-rule bypasses, HTTP/2 host confusion, and a stricter client_auth being silently dropped when TLS policies merge. Some fixes change behavior and may break apps that rely on the old behavior. Upgrade FrankenPHP to 1.13.5 and Caddy to 2.11.9.
Fixed in
FrankenPHP 1.13.5, Caddy 2.11.9
Latest updates (2)
2026-11-05: Caddy 2.11.9 ships, fixing a 2.11.8 regression and FastCGI security issues (GHSA-q34h-28rj-5hjm, GHSA-9cr8-wmhh-39pm and others) plus a TLS client_auth merge bug. (source)
2026-11-05: FrankenPHP 1.13.5 ships as a security release fixing 17 vulnerabilities, most found by Anthropic's OSS Scanner, and upgrades to Caddy 2.11.9 and Mercure 1.1.6. (source)
FreeRDP, the open-source RDP client, server and proxy library used by Remmina, GNOME Connections and many Linux desktops and thin clients, shipped 3.32.2 as a security release that hardens it against malicious client and server messages, with more than 40 GitHub advisories. CVE records for them were published on November 3 and 4. CVE-2026-121392 (CVSS 4.0 9.1) lets a man-in-the-middle holding a certificate from any client-trusted CA bypass server identity checks in tls_verify_certificate and intercept RDP sessions. CVE-2026-121953 (9.3) is a use-after-free in the client's rdpgfx channel that a malicious RDP server can trigger to crash the client or potentially execute code, and CVE-2026-121952 (9.3) is an out-of-bounds write in freerdp-proxy reachable by a connecting client. The rest are heap overflows, use-after-frees, a Kerberos double free and denial-of-service bugs, most needing the user to connect to a malicious or compromised server. Update FreeRDP to 3.32.2 or your distribution's patched package.
2026-11-04: CVE records CVE-2026-121950 to CVE-2026-121964 are published for heap overflows, use-after-frees and denial-of-service bugs fixed in 3.32.2. (source)
2026-11-03: CVE-2026-121392, a certificate identity validation bypass in tls_verify_certificate, is published. (source)
2026-10-12: FreeRDP releases 3.32.2 as a regression, bugfix and security release listing more than 40 GitHub security advisories. (source)
A command-injection flaw in the widely used npm package open (versions 7.2.0 to before 11.0.0) lets a crafted URL run PowerShell code on Windows and WSL: open placed the URL unescaped inside a PowerShell double-quoted string, where $(...) subexpressions execute. Tools that open server-supplied URLs are exposed. A malicious or hijacked remote MCP server can answer with an OAuth authorization endpoint containing a payload, and it runs as soon as the developer connects or signs in. Affected apps named in the CVE record are the OpenCode AI coding agent 1.0.137 to 1.18.32 (fixed in 1.18.33 on Sep 28 without an advisory) and MCP Inspector 2.0.0 to 2.11.0 (fixed in 2.12.0, GHSA-q978-78c7-phqw). open itself fixed the escaping in 11.0.0. VulnCheck assigned CVE-2026-119025 on Oct 29 2026 and rates it High (CVSS 4.0 7.7); Pillar Security, which found it, had proposed 9.8 for OpenCode. macOS and Linux are not affected; no in-the-wild exploitation reported.
Fixed in
open 11.0.0 (requires Node.js 20), OpenCode 1.18.33, @modelcontextprotocol/inspector 2.12.0
2026-10-29: VulnCheck assigns CVE-2026-119025 to the root cause in open 7.2.0 to before 11.0.0, listing OpenCode and MCP Inspector as affected; rated High (CVSS 4.0 7.7). (source)
2026-10-22: The Hacker News reports the flaw; upgrade to 1.18.33, run `opencode mcp list`, and remove unvetted remote MCP servers. (source)
2026-10-21: MCP Inspector publishes GHSA-q978-78c7-phqw (high): a remote MCP server can run code through the OAuth authorization URL on Windows and WSL; fixed in 2.12.0. (source)
ISC disclosed a remotely exploitable memory-corruption bug in BIND 9's named on Oct 21. An attacker who captures a GSS-TSIG signed UDP message (the Kerberos signing used for Active Directory-style dynamic DNS updates) can replay it while the negotiated key is still valid and, combined with crafted queries to a zone they control, may get remote code execution, or at least crash named. CVSS 8.1 (High). No workaround. Affects BIND 9.11.0 through 9.18.50, 9.20.0 through 9.20.29 and 9.21.0 through 9.21.26. Fixed in 9.20.30 and 9.21.27; ISC lists no 9.18 fix. ISC knows of no active exploitation. Debian trixie has a fixed package; Debian bookworm and Ubuntu are still waiting.
2026-10-22: Debian publishes DSA-6576-1 with a fixed bind9 for trixie; bookworm (9.18) is still listed as vulnerable. (source)
2026-10-22: Japan's registry JPRS issues an urgent alert telling BIND operators to upgrade. (source)
2026-10-21: ISC publicly discloses CVE-2026-94300 and ships BIND 9.20.30 and 9.21.27. ISC found it in internal testing; Dave Aitel of OpenAI reported it independently. (source)
SonicWall published SNWLID-2026-0014 covering multiple SonicOS flaws. Primary tracked here: CVE-2026-19868, a use-of-single-factor-authentication issue in the SonicOS SSLVPN service that may allow a user with valid credentials for an alternate authentication method to bypass SAML authentication controls and SAML-enforced MFA (CWE-308). CISA added it to the KEV catalog on 2026-11-06 (due 2026-11-20; ransomware use Unknown). NVD (published 2026-11-06T19:17:48Z by PSIRT@sonicwall.com) lists Gen7 SonicOS 7.3.3-7015 and older and Gen8 8.2.2-8015 and older as affected; no CVSS in NVD yet. Same advisory also covers sibling SonicOS CVEs CVE-2026-82308 (pre-auth SSL-VPN heap overflow DoS), CVE-2026-107154 / CVE-2026-107155 (management interface), CVE-2026-19703 (weak password hashing on virtual appliances), and CVE-2026-77708 / CVE-2026-77709 (pre-auth TACACS+ buffer overflows). Confirm fixed builds on the vendor advisory.
Fixed in
See SNWLID-2026-0014 for fixed SonicOS builds; NVD lists Gen7 newer than 7.3.3-7015 and Gen8 newer than 8.2.2-8015 as outside the affected range
2026-11-06: SonicWall PSIRT publishes SNWLID-2026-0014; NVD receives CVE-2026-19868 (SSLVPN SAML/MFA bypass via alternate auth) and sibling SonicOS CVEs CVE-2026-82308, CVE-2026-107154, CVE-2026-107155, CVE-2026-19703, CVE-2026-77708, CVE-2026-77709. Affected: Gen7 7.3.3-7015 and older; Gen8 8.2.2-8015 and older. (source)
2026-11-06: CISA adds CVE-2026-19868 to the Known Exploited Vulnerabilities Catalog (due 2026-11-20; known ransomware use Unknown). Required action: apply vendor mitigations per BOD 26-04. (source)
Erica Windisch of Cyberdione Labs published Tailworm, 11 lab-reproduced findings in Tailscale v1.104.2 and headscale v0.29.5 without coordinated disclosure. The main one: with Tailscale SSH enabled, unix-socket port forwarding runs as root, so anyone with one approved device and an SSH rule can reach root on the target and capture enrollment keys. Others cover experimental app connectors bypassing ACLs (cloud metadata, internal LANs), a tailscale serve forward to the local API, and headscale/DERP memory exhaustion. Tailscale confirmed the SSH issues as serious privilege-escalation bugs and on Oct 28 released Tailscale 1.104.4 with bulletins TS-2026-014 (SSH unix-socket forwarding to root-only sockets such as the Docker socket or local API), TS-2026-015 (SSH session memory exhaustion), TS-2026-016 (serve unix backends reaching privileged sockets) and TS-2026-017 (HTTP/2 serve capability mix-up). Upgrade to 1.104.4 or newer; the bulletins do not cover the app-connector or headscale findings. It is lateral movement from a foothold, not internet RCE.
Fixed in
tailscale 1.104.4, headscale 0.29.7 (control server; nodes still need Tailscale v1.104.4+)
Latest updates (3)
2026-10-30: Headscale, the open-source Tailscale control server, ships 0.29.7 as a security release fixing Tailworm findings and issues from Anthropic's oss-scanner (none critical, some can cause denial of service). It recommends upgrading to Headscale 0.29.7 and Tailscale v1.104.4 or newer, noting the SSH fixes run in tailscaled on each node, so upgrading Headscale alone does not apply them. (source)
2026-10-28: Tailscale releases v1.104.4 and publishes TS-2026-014 (Tailscale SSH unix-socket forwarding ran as root and could reach root-only sockets such as the Docker socket or the tailscaled local API), TS-2026-015 (SSH session memory exhaustion DoS), TS-2026-016 (serve unix backends could reach privileged sockets, including the local API) and TS-2026-017 (HTTP/2 serve could mix up Tailscale-App-Capabilities across mounts), all fixed in 1.104.4. Mitigation if you can't upgrade: TS_SSH_DISABLE_FORWARDING=1. (source)
2026-10-27: SecurityWeek reports the Tailworm report was updated on Oct 27 to say fixes for several findings have landed in Tailscale's public code repositories, but none is in a numbered release yet; Tailscale still targets a patched client by the end of Oct 27 and neither side reports exploitation in the wild. (source)
CVE-2026-83844 is an out-of-bounds read in Microsoft Paint that Microsoft rates Important with CVSS 8.1. It can allow remote code execution over a network with no sign-in and no user interaction, at high attack complexity. Microsoft published the issue on Oct 13 2026 as a Microsoft Store security update, not a Windows Update. The public CVE record still lists Paint as affected below 11.2605.91.0, while MSRC remediation names fixed build 11.2609.401.0 — scanners that trust only the CVE range may wrongly clear older Store builds. As of Nov 6 2026, The Hacker News and RFF reported that Store catalogs in multiple countries were serving 11.2609.401.0. Admins should inventory every user copy and the provisioned package and update Paint through the Store (or MDM Store-app update scans) to 11.2609.401.0 or later. Not known to be exploited or publicly disclosed at Microsoft's publication; not in CISA KEV when THN checked Nov 6.
Fixed in
Microsoft Paint 11.2609.401.0 or later via Microsoft Store (not Windows Update)
2026-11-06: The Hacker News and RFF report Microsoft Store now offering Paint 11.2609.401.0 in checked markets; Windows Update still does not deliver the fix. (source)
2026-11-05: RFF reports Store catalogs in multiple countries still topping out at Paint 11.2605.81.0, below the MSRC-named fixed build 11.2609.401.0. (source)
2026-10-13: Microsoft publishes CVE-2026-83844 for a Paint out-of-bounds read RCE delivered as a Store security update. (source)
FortiOS, FortiPAM, FortiProxy, and FortiSwitchManager contain a path traversal vulnerability that may allow an authenticated attacker with admin profile and at least read-write permissions to write or delete arbitrary files via specific CLI commands. Added to CISA KEV on 2026-10-16.
Gitea 28.3.0 (third maintenance release of the 28 series) ships a large security batch. CVE-2026-120263: when ACCOUNT_LINKING=auto and registration is enabled, the link-account sign-up form looked up the account to link by attacker-submitted username/email instead of the OAuth2 identity, allowing passwordless takeover of existing accounts. The release also fixes high-severity Actions issues (fork workflows obtaining secrets/write tokens via review commit_id; Actions task tokens authenticating to the REST API without scope), stored XSS via artifact Content-Encoding, write-collaborator release/milestone/label destruction on archived repos, and additional medium issues around token scopes. Gitea strongly recommends upgrading ASAP. Fixed in 28.3.0.
2026-11-01: Gitea publishes multiple security advisories for the 28.3.0 release, including GHSA-8p3x-w464-4g6v (OAuth2 account linking) and high-severity Actions/token issues. (source)
2026-11-01: Gitea blog: Release of 28.3.0 documents CVE-2026-120263 and related security fixes; strongly recommends upgrading. (source)
On Nov 6 2026 AWS published Important bulletin 2026-166-AWS for two memory-safety issues in the Babelfish TDS protocol layer used by Babelfish for Aurora PostgreSQL (SQL Server wire protocol). CVE-2026-119923: a memory-safety issue in the TDS connection-handshake handling could allow a remote, unauthenticated actor to affect the confidentiality, integrity, and availability of the database process. CVE-2026-119924: a memory-safety issue in the TDS remote-procedure-call request handling could allow a remote, authenticated actor to read adjacent process memory or cause a denial of service. AWS addressed the issues in Babelfish for PostgreSQL extension versions 4.10, 5.6 and 6.1, and lists patched Aurora PostgreSQL versions with Babelfish enabled. Workaround until upgraded: restrict network access to the TDS port (default 1433) to trusted sources using VPC security groups.
2026-11-06: AWS publishes Important bulletin 2026-166-AWS for CVE-2026-119923 (TDS connection-handshake memory-safety; remote unauthenticated actor can affect confidentiality, integrity, and availability of the database process) and CVE-2026-119924 (TDS RPC request handling memory-safety; remote authenticated actor can read adjacent process memory or cause DoS). Fixed in Babelfish extension 4.10 / 5.6 / 6.1 and listed Aurora PostgreSQL versions. Workaround: restrict TDS port 1433 via VPC security groups. (source)
On Nov 6 2026 AWS published Important bulletin 2026-173-AWS for CVE-2026-122193 (GHSA-95g5-m242-wvhr): an OS command injection issue in an example from the Trusted Remote Execution (REX) set-up documentation. The example script creates AWS Systems Manager documents Rex-RO and Rex-RW for running scripts on managed instances; due to how that example handled input parameters (including scriptArgs), a user permitted to invoke the SSM document could supply crafted input that runs unintended OS commands on the managed instance outside the REX sandbox. This is not a flaw in a shipped REX artifact — only operators who created those SSM documents by following the guide are affected. AWS removed the example from the guide. Mitigation: delete any Rex-RO / Rex-RW SSM documents created from the example in each account and Region where it was run. No workaround other than deleting the documents.
Fixed in
Example removed from REX set-up guide; delete any created Rex-RO / Rex-RW SSM documents
2026-11-06: AWS publishes Important bulletin 2026-173-AWS for CVE-2026-122193: OS command injection in REX set-up guide example SSM documents Rex-RO/Rex-RW; example removed from guide; delete created documents. (source)
Charmbracelet Wish (SSH server middleware for Go) before 2.0.5 has a path traversal in the git middleware that lets authenticated SSH users escape the configured repository directory with crafted repo paths. Fixed in v2.0.5 (commit 533db0d). CVSS 8.1 per NVD/VulnCheck.
Cisco published a high-severity advisory for UCS Central Software, its management platform for multiple Cisco Unified Computing System domains. CVE-2026-76506 (CVSS 7.6, CWE-22 path traversal) is insufficient validation of user-supplied input in an HTTP request to the web-based management interface. An authenticated remote attacker with valid user credentials can send a crafted request to execute arbitrary commands, write arbitrary files to the filesystem and obtain sensitive information. It affects UCS Central regardless of configuration, and there are no workarounds: upgrade release 2.1 to 2.1(2c), and migrate release 2.0 and earlier to a fixed release. Cisco PSIRT says it is not aware of public announcements or malicious use.
A flaw in APT, Debian's package manager, lets a man-in-the-middle attacker replace a package after it has been verified, resulting in arbitrary code execution as root. Per Debian's advisory DSA-6607-1, APT can merge two different package downloads after an HTTP redirect gives them the same URI, and improper file handling when resuming an HTTP download allows the swap. It was found by Max Justicz. Debian stable (trixie) is fixed in apt 3.0.4, which also addresses several out-of-bounds access issues, and Debian 12 (bookworm) is fixed in apt 2.6.2 (DLA-4880-1); Debian testing (forky) was still listed as vulnerable. Ubuntu published fixes on Nov 4 for 26.10, 26.04 LTS, 24.04 LTS and 22.04 LTS, with Ubuntu Pro fixes for older releases. The bug needs a non-zero Acquire::Retries setting, which is the default from Debian 12 and Ubuntu 22.04 on, so older releases are unaffected in a default configuration. No exploitation has been reported.
2026-11-04: Debian publishes DSA-6607-1 for apt (CVE-2026-109518): a man-in-the-middle attacker can replace a package after verification and run code as root; fixed in apt 3.0.4 for trixie. (source)
2026-11-04: The advisory reaches the Hacker News front page. (source)
2026-11-04: Ubuntu ships fixed apt for 26.10, 26.04 LTS, 24.04 LTS and 22.04 LTS (Ubuntu Pro fixes for older releases) and rates it CVSS 4.0 7.7 high; Debian's tracker lists bookworm fixed in apt 2.6.2 (DLA-4880-1), with forky still vulnerable. (source)
INFINITT Healthcare advisory INFINITT-SA-2026-001 (Nov 5 2026) and NCSC.ch CVE records cover two issues in the ULite medical imaging universal viewer. CVE-2026-19877 (High, CVSS 3.1 7.5) lets an unauthenticated remote attacker brute-force the low-entropy password on image-share links with no attempt limiting, gaining unauthorized access to shared medical images within the link validity window on ULite 5.0.2.1_BN48_20260528 and earlier. CVE-2026-19876 (Medium, CVSS 6.1) is a pre-authentication reflected XSS on the login/SSO error page via an unsanitized request parameter. Both are fixed in ULite 5.0.2.1 BN50. INFINITT says there is no indication of exploitation and is deploying the update to customers under existing maintenance agreements; sites should schedule the BN50 update with INFINITT support.
2026-11-06: NCSC.ch publishes CVE-2026-19877 (High, CVSS 7.5): unauthenticated brute-force of low-entropy image-share link passwords on ULite <= 5.0.2.1_BN48_20260528; and CVE-2026-19876 (Medium, CVSS 6.1): pre-auth reflected XSS on the login/SSO error page. Fixed in 5.0.2.1 BN50. (source)
2026-11-05: INFINITT Healthcare publishes INFINITT-SA-2026-001: security update for ULite share-link access control weakness and reflected XSS in the web interface; fixed builds BN50 deploying under maintenance agreements; no indication of exploitation. (source)
Google promoted Chrome 157 to the stable channel on Nov 3 2026 (157.0.8103.38 on Linux, 157.0.8103.38/.39 on Windows and Mac) with 213 security fixes. Seven are rated critical: buffer overflows in Dawn (CVE-2026-121156, CVE-2026-121247), memory corruption in ANGLE (CVE-2026-121229), and use-after-free bugs in WebGL (CVE-2026-121246), Prerender (CVE-2026-121035), Geometry (CVE-2026-121203) and Core (CVE-2026-121126). Many of the high-severity bugs were credited to AI-assisted research from Anthropic and OpenAI Codex Security. Google has not said any are exploited. The rollout is staged over days or weeks, so update and relaunch rather than waiting.
2026-11-04: Hong Kong's GovCERT issues an alert telling users to update Chrome to 157.0.8103.38 or later, warning the flaws could allow remote code execution. (source)
2026-11-03: Chrome 157 promoted to stable with 213 security fixes, seven of them critical (Dawn, ANGLE, WebGL, Prerender, Geometry, Core); none reported exploited. (source)
mark3labs/mcp-go's MCPServer.HandleMessage decodes JSON-RPC with encoding/json, which matches member names case-insensitively, folds ſ (U+017F) to s and K (U+212A) to k, and keeps the last of duplicate members. A request with Method/METHOD/paramſ, or with method given twice, still runs as the named call. mcp-go does not authorize calls itself, so this matters when a gateway or proxy in front allowlists method / params.name with an exact-match parser: the proxy may not see the call the server executes. The official Model Context Protocol Go SDK had the same class of bug (CVE-2026-27896 / GHSA-wvj2-96wp-fq3f) and fixed it in 1.3.1. Reported privately to contact@mark3labs.com on 2026-08-16 with no reply, then disclosed. Affected through v1.5.0 (tested on v1.0.0-beta.1, v1.4.0 and main after v1.5.0); no fix released. Workarounds: reject messages whose member names are not exactly the defined ones or that repeat a member, or re-encode before forwarding.
Latest updates (3)
2026-11-04: golang/vulndb#6911 opened describing the mcp-go JSON-RPC case/Unicode-folding proxy-allowlist bypass; fixed in: not fixed. (source)
2026-11-04: Public disclosure filed as mark3labs/mcp-go#1164 after private report and follow-up went unanswered. (source)
2026-08-16: Issue reported privately to contact@mark3labs.com; no reply (per public disclosure). (source)
nodejs/undici shipped coordinated security releases 7.30.2 and 8.12.4 on Nov 5 2026 fixing six advisories. High GHSA-mqg3-3v4r-cm32 / CVE-2026-116592 lets an originless Agent with compose(redirect(), cache()) and a shared cache poison or disclose across users because only the final redirect hop was checked, so intermediate hops could bypass the earlier CVE-2026-109829 fix (affected >=7.30.1 <7.30.2 and >=8.12.2 <8.12.4). Five Medium issues in the same release cover same-origin redirect cache disclosure (GHSA-7xg3-fphp-28hg / CVE-2026-120474), cross-origin redirect revalidation disclosure (GHSA-5rm8-gv32-r954 / CVE-2026-116630), array-of-tuples headers bypassing redirect credential stripping (GHSA-9v6h-8rgp-m69f / CVE-2026-120415), unhandled exception DoS during cache-interceptor response replay (GHSA-3g9m-rv3r-w5w2 / CVE-2026-120222), and Subresource Integrity bypass when integrity metadata uses non-space ASCII whitespace such as a tab (GHSA-r4pc-4jcp-7pqc / CVE-2026-120423). Upgrade to undici 7.30.2 or 8.12.4.
SonicWall disclosed two flaws in its end-of-life Global Management System (GMS) 9.5.2 and earlier and said no security update will be released. CVE-2026-109593 lets an unauthenticated attacker with network access read internal GMS files and expose sensitive information (CVSS 9.1 from CISA's ADP). CVE-2026-109594 is a post-authentication SQL injection that can let a high-privileged attacker access or modify data in the GMS database (CVSS 6.8). The flaws were reported by Pentest-Tools researchers. SonicWall urges customers to move to Network Security Manager (NSM); until then, keep GMS off untrusted networks.
2026-11-04: BleepingComputer reports SonicWall's warning about the unpatched GMS flaws. (source)
2026-11-03: SonicWall publishes SNWLID-2026-0019 for CVE-2026-109593 and CVE-2026-109594 in EOL GMS and says no remediation will be released; it recommends upgrading to NSM. (source)
WordPress released 7.1.4 on Oct 28 2026 as a security release and recommends updating sites immediately. It fixes a weakness in the WordPress upgrade process that can lead to remote code execution during the upgrade window (reported by Umut Bayram), a long list of stored XSS issues (crafted block markup, Privacy Data Removal Requests, the Interactivity API, Customizer selective refresh, the Categories, Playlist, image and comments-pagination blocks, and specially crafted user logins), a stored denial of service in the Interactivity API, unpublished Customizer changesets being discoverable by unauthenticated users, and Notes being readable by privileged users who can't edit the post. It also updates the bundled root certificates because of the eMudhra revocation. Sites with automatic background updates get it automatically; the fixes are being backported to every branch still eligible for security fixes (back to 4.7), and those backports ship as they become ready. WordPress 7.2 Beta 3 has the same fixes. Wordfence rates the upgrade-window RCE CVSS 8.1 (unauthenticated, high complexity: WordPress 4.7 to 7.1.3 unpack full-package updates into a predictable, web-reachable folder under wp-content/upgrade, and an attacker who requests the staged installer in the seconds-long window can reach PHP code execution). Wordfence says attackers began trying to exploit it within minutes of its firewall rule going live on Oct 28 and logged more than 18,500 requests to staged setup-config.php paths in 21 hours; it reports attempts, not confirmed compromises.
Fixed in
7.1.4, 7.0.8, 6.9.11, backports down to 4.7.39, 7.2 Beta 3
2026-11-06: Wordfence publishes that attackers are actively probing for the WordPress upgrade-window RCE (CVE-2026-119977); discussed on Hacker News. Sites should stay on patched branches and keep wp-content/upgrade non-web-executable during upgrades. (source)
2026-11-06: The Hacker News reports Picus Security researcher Umut Bayram warning that installing the 7.1.4 (or branch) fix can still expose some sites one last time: the update that installs the fix unpacks using the old updater, so complete-package upgrades on 5.9+ can leave a predictable installer under wp-content/upgrade during the window. He advises denying web access to wp-content/upgrade before updating (or doing a manual out-of-web-root update); partial same-branch packages may not include installer files, but WordPress can fall back to a complete package. (source)
2026-11-03: CISA adds CVE-2026-119977 (WordPress Core TOCTOU race during full-package upgrades) to the Known Exploited Vulnerabilities catalog; federal agencies have until Nov 6 2026 to remediate. (source)
ZITADEL published three account takeover advisories for its Login V2 UI, with CVE records on Oct 29 2026. CVE-2026-119127 (CVSS 4.0 8.6) is a host header injection: unauthenticated attackers who know a victim's login name can poison password-reset, invite and email-verification links via the x-zitadel-public-host, x-zitadel-forward-host or x-forwarded-host headers, harvest the single-use codes and take over accounts, including instance administrators. CVE-2026-119128 (CVSS 4.0 9.1) lets unauthenticated attackers compute a valid verificationCheck cookie from the user ID and browser fingerprint and set the first credential (password, passkey or linked identity provider) of users that have no authentication method yet. CVE-2026-119129 (CVSS 4.0 8.6) is missing authorization in OAuth 2.0 device authorization: a signed-in user who opens a link carrying the attacker's user code hands over their tokens, and project access settings can be bypassed via the OIDC device authorization API. 4.0.0 before 4.19.5 is affected by all three; 3.0.0 through 3.4.15 is also affected by the device authorization bug.
iTerm2, the popular macOS terminal, has an argument injection bug in its x-man-page: URL handler in versions before 3.7.7beta2. Per the CVE record, percent-encoded NUL characters survive iTerm2s input sanitization and are later turned into whitespace, so a crafted link can inject extra arguments into /usr/bin/man, including -P to pick an attacker-controlled pager command that runs with the users privileges. The victim has to open the crafted URL in iTerm2, for example by clicking it in a browser where iTerm2 handles the x-man-page scheme (a browser or macOS confirmation may appear) or in terminal output. The record notes it exists because of an incomplete fix for CVE-2023-46321. The fix landed in 3.7.7beta2, and stable iTerm2 3.7.7 (built Nov 3 2026) now checks man section and page names against a character allowlist; the release notes credit vin01 for finding it. The CVE record rates it CVSS 4.0 8.7 (high). No exploitation has been reported.
2026-11-05: CVE-2026-123014 is published: percent-encoded NUL characters let a crafted x-man-page: URL inject arguments such as -P into /usr/bin/man and run arbitrary commands, an incomplete fix for CVE-2023-46321; CVSS 4.0 8.7. (source)
2026-11-03: iTerm2 3.7.7 ships with a hardened x-man-page: handler: man section and page names are checked against an allowlist and URLs with disallowed characters are rejected; the release notes credit vin01 for finding the bug. (source)
Sansec reports a criminal group that has injected a JavaScript loader into the cms_block table of 953 online stores since July 2025, mostly Magento. The newest loader domain, deepveindrilling.com (registered Oct 6, 2026), was found on 83 stores, and since Oct 23 it fingerprints visitors so scanners get nothing. Sansec ties initial access to PolyShell (CVE-2026-48356, fixed in Adobe APSB26-73) and SessionReaper (CVE-2025-54236, APSB25-88), plus rogue admin accounts on fully patched stores. Store owners should patch, audit cms_block and cms_page, remove rogue admins and rotate credentials.
Fixed in
APSB26-73 (2.4.x-2026-jul security releases) for CVE-2026-48356, APSB25-88 for CVE-2025-54236
2026-10-27: CISA adds CVE-2026-48356 (Adobe Commerce and Magento unrestricted file upload, PolyShell) to its Known Exploited Vulnerabilities catalog based on evidence of active exploitation. (source)
2026-10-24: Sansec publishes its research: 953 stores across five loader domains since July 2025, 83 on deepveindrilling.com (60 of them earlier victims). (source)
2026-10-23: Operator rewrites the blocks with a fingerprinting loader that hides from automated browsers; 9 more stores infected, including a fully patched 2.4.7-p10 store with a rogue admin account. (source)
