OS Command Injection

(1) boardData102.php, (2) boardData103.php, (3) boardDataJP.php, (4) boardDataNA.php, and (5) boardDataWW.php in Netgear WN604 before 3.3.3 and WN802Tv2, WNAP210v2, WNAP320, WNDAP350, WNDAP360, and. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

In SolarWinds Log & Event Manager (LEM) before 6.3.1 Hotfix 4, a menu system is encountered when the SSH service is accessed with "cmc" and "password" (the default username and password). Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and EPSS exploitation probability 49.9%.

A command injection vulnerability was discovered on the Zyxel EMG2926 home router with firmware V1.00(AAQT.4)b8. Rated high severity (CVSS 8.8), this vulnerability is remotely exploitable, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

An issue was discovered in Trend Micro InterScan Messaging Security (Virtual Appliance) 9.1-1600. Rated high severity (CVSS 8.8), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 64.6%.

dnslookup.cgi on NETGEAR DGN2200 devices with firmware through 10.0.0.50 allows remote authenticated users to execute arbitrary OS commands via shell metacharacters in the host_name field of an HTTP. Rated high severity (CVSS 8.8), this vulnerability is remotely exploitable, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

ping.cgi on NETGEAR DGN2200 devices with firmware through 10.0.0.50 allows remote authenticated users to execute arbitrary OS commands via shell metacharacters in the ping_IPAddr field of an HTTP. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

Unauthenticated Remote Command injection as root occurs in the Western Digital MyCloud NAS 2.11.142 /web/google_analytics.php URL via a modified arg parameter in the POST data. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and EPSS exploitation probability 91.8%.

The isMail transport in PHPMailer before 5.2.20 might allow remote attackers to pass extra parameters to the mail command and consequently execute arbitrary code by leveraging improper interaction. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and EPSS exploitation probability 93.4%.

Cisco Adaptive Security Appliance (ASA) Software before 8.4(1) on ASA 5500, ASA 5500-X, PIX, and FWSM devices allows local users to gain privileges via invalid CLI commands, aka Bug ID CSCtu74257 or. Rated high severity (CVSS 7.8), this vulnerability is low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

Apache James Server 2.3.2, when configured with file-based user repositories, allows attackers to execute arbitrary system commands via unspecified vectors. Rated high severity (CVSS 8.1), this vulnerability is remotely exploitable, no authentication required. Public exploit code available and EPSS exploitation probability 74.9%.

Apache Struts 2.3.19 to 2.3.20.2, 2.3.21 to 2.3.24.1, and 2.3.25 to 2.3.28, when Dynamic Method Invocation is enabled, allow remote attackers to execute arbitrary code via method: prefix, related to. Rated high severity (CVSS 8.1), this vulnerability is remotely exploitable, no authentication required. Public exploit code available and EPSS exploitation probability 94.0%.

xymond in Xymon 4.1.x, 4.2.x, and 4.3.x before 4.3.25 allow remote authenticated users to execute arbitrary commands via shell metacharacters in the adduser_name argument in (1) web/useradm.c or (2). Rated high severity (CVSS 8.8), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 56.3%.

Endian Firewall before 3.0 allows remote attackers to execute arbitrary commands via shell metacharacters in the (1) NEW_PASSWORD_1 or (2) NEW_PASSWORD_2 parameter to cgi-bin/chpasswd.cgi. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 86.7%.

Watchguard XCS 9.2 and 10.0 before build 150522 allow remote authenticated users to execute arbitrary commands via shell metacharacters in the id parameter to ADMIN/mailqueue.spl. Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 72.5%.

AVM Fritz!Box allows remote attackers to execute arbitrary commands via shell metacharacters in the var:lang parameter to cgi-bin/webcm. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 87.8%.

The cpanel function in go_site.php in GoAutoDial GoAdmin CE before 3.3-1421902800 allows remote attackers to execute arbitrary commands via the $type portion of the PATH_INFO. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 87.5%.

The saveObject function in moadmin.php in phpMoAdmin 1.1.2 allows remote attackers to execute arbitrary commands via shell metacharacters in the object parameter. Rated high severity (CVSS 7.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 88.1%.

The D-Link DIR-645 Wired/Wireless Router Rev. Rated high severity (CVSS 8.8), this vulnerability is no authentication required, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

The management console on the Symantec Web Gateway (SWG) appliance before 5.2.2 allows remote authenticated users to execute arbitrary OS commands by injecting command strings into unspecified PHP. Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 74.0%.

The fetch_url function in usr.bin/ftp/fetch.c in tnftp, as used in NetBSD 5.1 through 5.1.4, 5.2 through 5.2.2, 6.0 through 6.0.6, and 6.1 through 6.1.5 allows remote attackers to execute arbitrary. Rated high severity (CVSS 7.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 85.0%.

GNU Bash through 4.3 bash43-026 does not properly parse function definitions in the values of environment variables, which allows remote attackers to execute arbitrary commands via a crafted. Rated high severity (CVSS 8.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

GNU Bash through 4.3 bash43-025 processes trailing strings after certain malformed function definitions in the values of environment variables, which allows remote attackers to write to files or. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

GNU Bash through 4.3 processes trailing strings after function definitions in the values of environment variables, which allows remote attackers to execute arbitrary code via a crafted environment,. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

The network interface configuration page (netinterface) in Sophos Web Appliance before 3.8.2 allows remote administrators to execute arbitrary commands via shell metacharacters in the address. Rated high severity (CVSS 8.5), this vulnerability is remotely exploitable. Public exploit code available and EPSS exploitation probability 75.7%.

The Cisco WAP4410N access point with firmware through 2.0.6.1, WRVS4400N router with firmware 1.x through 1.1.13 and 2.x through 2.0.2.1, and RVS4000 router with firmware through 2.0.3.2 allow remote. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 65.6%.

cgi-bin/admin/servetest in TP-Link IP Cameras TL-SC3130, TL-SC3130G, TL-SC3171, TL-SC3171G, and possibly other models before beta firmware LM.1.6.18P12_sign6 allows remote attackers to execute. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 72.3%.

Directory traversal vulnerability in processImageSave.jsp in DCNM-SAN Server in Cisco Prime Data Center Network Manager (DCNM) before 6.2(1) allows remote attackers to write arbitrary files via the. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 88.7%.

The close_connections function in /opt/cma/bin/clear_keys.pl in Sophos Web Appliance before 3.7.9.1 and 3.8 before 3.8.1.1 allows local users to gain privileges via shell metacharacters in the second. Rated high severity (CVSS 7.2), this vulnerability is low attack complexity. Public exploit code available and no vendor patch available.

The get_referers function in /opt/ws/bin/sblistpack in Sophos Web Appliance before 3.7.9.1 and 3.8 before 3.8.1.1 allows remote attackers to execute arbitrary commands via shell metacharacters in the. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 92.7%.

ginkgosnmp.inc in HP System Management Homepage (SMH) allows remote authenticated users to execute arbitrary commands via shell metacharacters in the PATH_INFO to smhutil/snmpchp.php.en. Rated critical severity (CVSS 9.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 46.3%.

The NetWorker command processor in rrobotd.exe in the Device Manager in EMC AlphaStor 4.0 before build 800 allows remote attackers to execute arbitrary commands via a DCP "run command" operation. Rated critical severity (CVSS 9.3), this vulnerability is remotely exploitable. Public exploit code available and EPSS exploitation probability 78.7%.

Mutiny Standard before 4.5-1.12 allows remote attackers to execute arbitrary commands via the network-interface menu, related to a "command injection vulnerability.". Rated high severity (CVSS 8.5), this vulnerability is remotely exploitable. Public exploit code available and EPSS exploitation probability 70.6%.

lhn/public/network/ping in HP SAN/iQ before 9.5 on the HP Virtual SAN Appliance allows remote authenticated users to execute arbitrary commands via shell metacharacters in the second parameter. Rated high severity (CVSS 7.7), this vulnerability is low attack complexity. Public exploit code available and EPSS exploitation probability 44.8%.

The web browser plugin for Ubisoft Uplay PC before 2.0.4 allows remote attackers to execute arbitrary programs via the -orbit_exe_path command line argument. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 80.2%.

The management console in Symantec Web Gateway 5.0.x before 5.0.3.18 allows remote attackers to execute arbitrary commands via crafted input to application scripts. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 83.4%.

An ActiveX control in KeyHelp.ocx in KeyWorks KeyHelp Module (aka the HTML Help component), as used in GE Intelligent Platforms Proficy Historian 3.1, 3.5, 4.0, and 4.5; Proficy HMI/SCADA iFIX 5.0. Rated critical severity (CVSS 9.3), this vulnerability is remotely exploitable. Public exploit code available and EPSS exploitation probability 69.4%.

sapi/cgi/cgi_main.c in PHP before 5.3.12 and 5.4.x before 5.4.2, when configured as a CGI script (aka php-cgi), does not properly handle query strings that lack an = (equals sign) character, which. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Actively exploited in the wild (cisa kev) and public exploit code available.

Remote code execution in Windows Help and Support Center (helpctr.exe) on Windows XP and Server 2003 allows unauthenticated attackers to bypass URL validation and execute arbitrary commands via crafted hcp:// URLs. The MPC::HexToNum function fails to properly validate escape sequences, enabling attackers to circumvent the trusted documents whitelist. EPSS indicates 92.20% exploitation probability with publicly available exploit code. Microsoft addressed this via MS10-042 in July 2010 after public disclosure in June 2010.

OS command injection in Cacti network monitoring (fixed in 1.2.31) lets an authenticated administrative user inject shell metacharacters into a Data Input Method's input string, which Cacti passes to the shell when running data-collection scripts, yielding code execution on the poller host. The 1.2.31 hardening (advisory GHSA-xq98-376r-hv9j) also closes CSV/formula injection in host and automation exports, a content-file path traversal in index.php, and unsafe rrdtool argument handling. No public exploit identified at time of analysis and the issue is not listed in CISA KEV.

Command injection (CWE-78) in MariaDB Server affects multiple branches (10.6.x, 10.11.x, 11.4.x, 11.8.x, and 12.3.x) and allows an authenticated high-privileged user to achieve full compromise of confidentiality, integrity, and availability over the network. Alpine Linux has shipped a fixed package (mariadb 11.8.8-r0), and an upstream advisory (GHSA-7v3p-h23x-8hwv) plus MDEV-39676 track the issue, but no public exploit identified at time of analysis and EPSS is very low at 0.06%.

OS command injection in MariaDB Server (CWE-78) affects multiple supported branches (10.6.x, 10.11.x, 11.4.x, 11.8.x, 12.3.x) and allows an attacker with high database privileges to execute arbitrary operating system commands on the database host, leading to full confidentiality, integrity, and availability compromise. The flaw is tracked by Alpine, the upstream MariaDB GHSA-rpgv-q6gv-684r advisory, and MDEV-39648, with patches already released across branches. No public exploit is identified at time of analysis and EPSS is very low (0.07%, 22th percentile), so the issue is currently treated as a high-severity patch-now rather than an in-the-wild emergency.

Command injection in Samba (CWE-78) carries a critical CVSS 9.8 (AV:N/AC:L/PR:N/UI:N) and is fixed in Alpine Linux package samba 4.23.8-r0, with Red Hat shipping errata RHSA-2026:22644, 22963, 25049, and 25979 for Enterprise Linux 6/7/9 and OpenShift Container Platform 4.0. The flaw allows remote attackers to inject operating-system commands against vulnerable Samba deployments, yielding full confidentiality, integrity, and availability impact. EPSS is modest at 0.23% (46th percentile) and no public exploit identified at time of analysis, but the unauthenticated network attack surface of Samba makes urgent patching warranted.

OS command injection in MariaDB Server (fixed in Alpine package mariadb 11.8.7-r0) allows a high-privileged authenticated database user to break the trust boundary between SQL and the underlying operating system, resulting in confidentiality, integrity, and availability impact across components (scope changed). No public exploit identified at time of analysis, and EPSS exploitation probability is very low (0.07%, 22nd percentile). Tracked upstream as MariaDB MDEV-39413 with security advisory GHSA-vwf7-w26c-9w5h.

OS command injection in Apache Kylin 4.0.4 through 5.0.4 allows an attacker who can influence the Spark job launch path to execute arbitrary operating system commands on the Kylin server, with the assessed CVSS vector indicating authenticated access (PR:L) is required. The vendor notes the impact is concentrated on Kylin services directly exposed to the public internet, while in-house, network-isolated deployments are mostly unaffected. No public exploit code, EPSS score, or CISA KEV entry was identified at time of analysis; upgrading to Kylin 5.0.5 resolves the issue.

OS command injection in Netgate pfSense CE 2.5.0 through 2.9.0 allows an authenticated OpenVPN client to run arbitrary shell commands as the low-privileged 'unbound' user by embedding shell metacharacters in the certificate common name or username that openvpn.learn-address.sh interpolates unquoted into a 'su -c' string. Exploitation requires three conditions to line up: valid VPN authentication (the CVSS vector is PR:L, so authenticated), the non-default 'DNS Resolver (unbound) client registration' feature enabled, and attacker-controlled CN or username - which is why the assessed severity is low (CVSS 2.3 in the v4 vector supplied; equivalent CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:L per independent assessment). Publicly available exploit code exists; there is no confirmation of active exploitation (not in CISA KEV) and impact is confined to the unbound user rather than root, so this is a genuine but low-priority issue.

QNAP Notes Station 3 versions prior to 3.9.11 contain an OS command injection flaw (CWE-78) that permits unauthenticated network attackers to run arbitrary commands on a NAS where the optional Notes Station 3 add-on is installed and reachable. The vendor's CVSS 4.0 vector (6.9) indicates low confidentiality, integrity, and availability impact, suggesting the injected commands execute in a constrained context rather than granting full system control. No public exploit code has been identified at time of analysis, and the primary exposure is limited to internet-facing or broadly accessible NAS instances running the vulnerable add-on.

Command injection in the OpenSSH client allows shell metacharacters placed in a command-line username - specifically '$' or a non-trailing backslash - to reach a shell when the ssh configuration expands the %u token inside a ProxyCommand or Match exec directive, giving arbitrary command execution with the privileges of the account that runs ssh. Reachability is conditional: a default OpenSSH configuration is not affected, and the attack requires a non-default config using %u in those directives plus a username that an attacker can control or influence through a script, wrapper, or application that forwards untrusted input as the username, with a user or process actually initiating the connection (UI:R). Affected builds are OpenSSH as shipped in Red Hat Enterprise Linux 6 through 10, Red Hat Hardened Images, and Red Hat OpenShift Container Platform 4; no public exploit identified at time of analysis. Note the scoring divergence: Red Hat assigns 5.4 (C:L/I:L/A:N), while our independent assessment is higher at CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:H because command execution under the invoking account can fully compromise confidentiality, integrity, and availability of that account's data and sessions.

AUTOM8N (WHM/cPanel plugin) through commit 82821f7 contains an OS Command Injection in borg_restore.cgi.

Arbitrary OS command execution on managed remote endpoints is possible in Palo Alto Networks Cortex XSOAR and Cortex XSIAM deployments that use the OS Query content pack, where an authenticated administrator can exploit command injection in the pack's command-building logic. Affected environments are limited to those with the OS Query content pack installed and operational; exploitation requires high administrative privileges and no user interaction. No public exploit has been identified at time of analysis, and the risk is assessed as legitimate but not top-tier priority given the privileged access prerequisite.

Unauthenticated OS command injection in WeGIA prior to 3.8.6 lets remote attackers execute arbitrary commands as the web-server user through the pre-installation web/instalador/instalador.php endpoint, which remains reachable whenever ../config.php has not yet been created. The attacker submits the Linux reinstall flow with a crafted 'usuario' POST value containing shell metacharacters that is interpolated into a shell_exec() mysql import command; the same endpoint also writes attacker-controlled values into an executable config.php and concatenates the 'nomebd' database name into a CREATE DATABASE statement. Exploitation is gated by a deployment-state precondition - the installer only processes POST data while config.php is absent, and the command sink is reached only on Linux - which the CVSS:3.1 vector's AC:H reflects; no confirmed active exploitation (CISA KEV) and no public exploit code identified at time of analysis, and authentication requirements are unauthenticated per the PR:N vector.

Command injection in Palo Alto Networks PAN-OS allows an authenticated administrator who can reach the management interface to bypass system restrictions and execute arbitrary commands as root on PA-Series and VM-Series firewalls and on Panorama (virtual and M-Series). Because exploitation requires valid administrator credentials plus network reachability to the management plane, real-world risk is materially reduced by following Palo Alto's best practice of restricting management-interface access to trusted internal IP addresses; Cloud NGFW and Prisma Access are explicitly not affected. No public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis, and the vendor-assigned CVSS 4.0 base score is 6.1 (Urgency: Amber), making this a moderate rather than top-tier priority.

Public-key pinning enforcement in curl/libcurl built against the wolfSSL TLS backend can be bypassed, affecting versions up to and including 8.22.0, so a man-in-the-middle attacker can present an alternate server certificate that the opt-in CURLOPT_PINNEDPUBLICKEY (--pinnedpubkey) check should have rejected. The weakness is heavily constrained: the library must be compiled with wolfSSL, the calling application must have explicitly enabled pinning (a non-default hardening feature), and the attacker must already hold a network man-in-the-middle position - the flaw nullifies an existing detection layer rather than creating interception capability. No public exploit code or CISA KEV entry was identified at time of analysis and no EPSS score was supplied, so risk is concentrated in the specific wolfSSL-plus-pinning deployment profile; the issue is fixed in curl 8.23.0.

Transfer-direction confusion in curl and libcurl's TFTP handling lets a malicious or misbehaving TFTP server cause a client to mishandle a read (download) transfer as a write (upload) or vice versa, affecting releases from 7.15.0 through 8.22.0. The curl project rates this LOW severity, and the independent assessment scores it CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N - high attack complexity, unauthenticated per the vector's PR:N, with only limited confidentiality impact and no integrity or availability impact. The fix ships in curl 8.23.0; there is no public exploit identified at time of analysis and no evidence of active exploitation, and because reaching the vulnerable code requires the application to actually perform a tftp:// transfer against an attacker-influenced server, the practical risk is low despite curl's ubiquity.

CRLF-driven IMAP command injection in curl/libcurl's IMAP APPEND upload path allows an attacker who controls part of the appended message body to smuggle protocol commands into a live IMAP session when CRLF conversion is explicitly enabled (curl --crlf or CURLOPT_CRLF). Affected releases are curl and libcurl up to and including 8.22.0; the fix ships in curl 8.23.0, released alongside the project's quarterly batch of twenty-four advisories. The curl project rates this one LOW severity, the independently assessed vector is CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N (network reachable, no privileges required by the attacker, high attack complexity, limited integrity impact), and no public exploit identified at time of analysis.

Credential leakage in curl 8.21.0 through 8.22.0 allows unauthenticated remote attackers to obtain Negotiate (SPNEGO/Kerberos) authentication tokens when a redirect or cross-origin transition causes the 'Authorization: Negotiate' header to be sent to a foreign origin. The flaw is rated LOW by the curl project and has an assessed CVSS 3.1 score of AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N; no public exploit code was identified at time of analysis. Exploitation requires the user to explicitly enable Negotiate authentication (e.g., via --negotiate or CURLAUTH_NEGOTIATE) and for the request flow to reach a foreign origin, limiting impact to token disclosure rather than code execution or service disruption.

Local file overwrite in the curl command-line tool (versions 7.20.0 through 8.22.0) occurs when a user runs a transfer with both the -J/--remote-header-name option and --no-clobber, and the connected HTTP server - or a man-in-the-middle on the connection - sends a 103 Early Hints interim response containing a Content-Disposition header. Because the early-hints header is honored for filename selection before the no-clobber protection is applied, an attacker-influenced filename can cause curl to overwrite a local file that --no-clobber was explicitly intended to protect, with impact limited to integrity of that file (no confidentiality or availability effect). The curl project rates this LOW, exploitation requires the victim to actively choose a specific non-default option combination, and no public exploit code or CISA KEV entry is identified at time of analysis; the flaw is fixed in curl 8.23.0.

Credential mixup during TLS connection reuse in curl/libcurl up to and including 8.22.0 can cause an application relying on automatic client-certificate selection to present the wrong client certificate on a pooled connection, exposing a transfer's identity context to the wrong destination or policy. Exploitation is not reachable by an arbitrary external attacker with a single request: it requires a libcurl-based host application that enables automatic client-certificate selection on a TLS backend that supports it and also reuses connections across transfers with differing certificate expectations, which is why the curl project rates it LOW (assessed CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N, base 3.7, confidentiality-only impact). The issue is one of twenty-four advisories published with curl 8.23.0; no public exploit code has been identified and the flaw is not confirmed actively exploited (not in CISA KEV) at time of analysis.

Credential disclosure in curl and libcurl at or below 8.22.0 causes origin host credentials to be sent to a configured HTTP proxy instead of being reserved for the intended origin server. The flaw only manifests when a deployment both routes transfers through an HTTP proxy and supplies credentials intended for the origin host, which is why the curl project rates it LOW severity and why the independent assessment scores it CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N - confidentiality-only impact with high attack complexity and no integrity or availability effect. There is no KEV listing and no public exploit identified at time of analysis, so real-world urgency is driven mainly by whether an environment uses untrusted or shared proxies alongside origin credentials. The fix ships in curl 8.23.0.

Persisted HSTS entries held by curl and libcurl can be erased when a malicious or on-path server returns a crafted Strict-Transport-Security response, allowing an attacker to strip previously stored HTTPS-only policy for other hosts. Only builds running up to and including 8.22.0 that have HSTS persistence explicitly enabled (the curl tool's --hsts <file> option, or libcurl with CURLOPT_HSTS pointed at a cache file) are exposed - this is not a default configuration, which is the main limiting factor. The assessed impact is integrity-only (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N), with no code execution or direct data disclosure; harm is indirect and deferred, since a later connection to a de-pinned host is only downgradable if the attacker can also intercept that subsequent traffic. No public exploit code has been identified at time of analysis and there is no indication of active exploitation. Fixes ship in curl 8.23.0, which resolves this issue alongside twenty-three other CVEs published on the same date.

Persisted HSTS state in curl/libcurl can be silently truncated when a previously exported HSTS cache is re-imported, because the import logic drops a longer-lived child entry (for example an includeSubDomains rule) in favor of a shorter-lived parent. Applications that explicitly enable curl's HSTS support via CURLOPT_HSTS or the --hsts option and persist that cache to disk across runs are affected; default curl usage with HSTS off is not. The resulting loss of HSTS enforcement only becomes exploitable if an on-path (man-in-the-middle) attacker is positioned to force a protocol downgrade to plaintext HTTP, so the curl project rates this severity LOW and no public exploit code or KEV/EPSS signal is present in the available data. Remote attackers, no authentication required per the assessed vector (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:N), but attack complexity is high and confidentiality/integrity impact is limited.

Cross-origin reuse of a Negotiate-authenticated connection in curl and libcurl (through 8.22.0) allows a connection authenticated to one origin to be reused for a request to a different origin when a single easy handle or connection cache serves multiple hosts. The curl project rates the issue LOW, and the assessed vector CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N shows no privileges required (PR:N) but a high attack complexity, with only a low-confidentiality impact and no integrity or availability effect - there is no memory corruption or code-execution component. Exploitation is conditional on the consuming application being built with Negotiate/SPNEGO (GSSAPI/Kerberos) support and deliberately reusing connections across more than one origin, so it is not a remotely triggerable flaw against a default curl configuration; no public exploit code or CISA KEV entry is identified at time of analysis, and no EPSS score was provided in the available data.

A crafted Set-Cookie header returned by a server can abort the loading of curl/libcurl's cookie jar in versions 8.7.0 through 8.22.0, disrupting the client's cookie-persistence routine instead of completing the jar load. The flaw is only reachable when the client is configured to use a persistent cookie jar (the -b/--cookie or -c/--cookie-jar options, or CURLOPT_COOKIEFILE/CURLOPT_COOKIEJAR in libcurl) and the attacker controls or has compromised the server end of the connection, so it is a network-reachable but low-impact integrity issue (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N). The curl project rates this LOW severity as part of its 8.23.0 release, no public exploit code or confirmed active exploitation has been identified at time of analysis, and no EPSS or KEV signal is associated with the CVE.

Password disclosure in curl and libcurl 7.62.0 through 8.22.0 occurs when an application fetches CURLINFO_EFFECTIVE_URL for a transfer whose URL carried embedded userinfo credentials (user:password@host): the effective-URL string returned by libcurl is not scrubbed of the password, so the secret flows into whatever sink the application controls - log files, console/UI output, telemetry, or downstream services. The curl project classifies this as LOW, and our independent assessment agrees (CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:N), since there is no remote attack vector and exposure is confined to a sink the host application already owns; no public exploit code or confirmed active exploitation was identified at time of analysis. Applications that never embed credentials in URLs, or that never read the CURLINFO_EFFECTIVE_URL field, are not affected.

FTP wildcard downloads in libcurl 7.21.0 through 8.22.0 can misroute transferred file data to the wrong application-supplied write callback, giving limited confidentiality and integrity impact (CVSS 3.1 AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:N) to unauthenticated remote attackers. The flaw is only reachable when a consuming application explicitly enables CURLOPT_WILDCARDMATCH with a glob-expanded FTP URL - a non-default option used by a small minority of integrations - and when the client fetches multiple matching files from an FTP server whose responses the attacker can influence or control; a default curl/libcurl build that does not use FTP wildcards is not exposed. The curl project rates this as LOW severity, the fix ships in curl/libcurl 8.23.0, and no public exploit has been identified at time of analysis; no EPSS score, KEV listing, or public proof-of-concept is present in the available data.

FTP and FTPS transfers performed by curl/libcurl through 8.22.0 can reuse a connection that carries a client certificate configured for an earlier, differently-configured transfer, so a later transfer may present the wrong client certificate to a server. It is a client-side connection-reuse logic error, not a server-triggerable flaw: it requires an application built on libcurl that sets CURLOPT_SSLCERT (or equivalent) for FTP(S) and reuses connections across multiple transfers on the same handle, and the assessed vector is CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N. The curl project rates this LOW, and it is one of twenty-four flaws disclosed alongside curl 8.23.0; no public exploit code has been identified and there is no confirmed active exploitation (CISA KEV) at time of analysis. Practical risk is limited to deployments that mix client-certificate and non-client-certificate FTP(S) transfers on shared handles or shared connection caches.

In curl and libcurl through 8.22.0, libcurl's SSH host-key verification callback (CURLOPT_SSH_KEYFUNCTION) can misreport a server host key that does not match the expected key as merely 'missing' from the known-hosts store, so applications that treat unknown keys more leniently than changed keys may accept an attacker-substituted key. The practical impact is weakened host authentication (limited confidentiality and integrity loss, CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:N), and exploitation requires an active man-in-the-middle on the SCP/SFTP path, a configured known-hosts file, and a callback whose policy distinguishes MISSING from MISMATCH. No CVSS vector, EPSS score, or KEV entry accompanies the public data and no public exploit code has been identified at time of analysis; the curl project itself rates this issue LOW, and it is one of twenty-four advisories fixed in the curl 8.23.0 release (only CVE-2026-92392, an HTTP/2 upload use-after-free, is rated HIGH).

Negotiate (Kerberos/SPNEGO) credentials are forwarded to the wrong origin by curl and libcurl 8.21.0 through 8.22.0 when the initially contacted server returns an HTTP redirect to a different host while redirect following is enabled. An attacker who controls or influences either the redirect response or the redirect destination can capture the Negotiate authentication token, resulting in limited confidentiality exposure rather than code execution or data modification. The flaw is rated LOW by the curl project, no public exploit code has been identified at time of analysis, and it was fixed in the curl 8.23.0 release published on 2026-10-14 alongside twenty-three other advisories.

A server-controlled response can erase the contents of a libcurl application's persisted cookie jar in curl/libcurl 8.18.0 through 8.22.0, allowing a malicious, compromised, or impersonated HTTP server to destroy stored cookie state. The flaw only triggers when the application has the cookie engine enabled together with cookie-jar file persistence (the curl tool's -c/--cookie-jar option or the CURLOPT_COOKIEJAR libcurl option); without that configuration the code path is unreachable. Impact is bounded to loss of stored cookie data - an integrity-only concern with no code execution, no credential disclosure, and no further system compromise - and no public exploit code was identified at time of analysis, which is consistent with the LOW severity assigned by the curl project and the AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N assessment.

OS command injection in Dell Unisphere for PowerMax vApp 9.2.4.x lets a remote attacker who already holds a high-privileged (administrative) account on the appliance run arbitrary operating-system commands on the underlying host. Because the CVSS vector requires PR:H and AC:H, exploitation is gated on prior admin-level authentication plus network reachability of the Unisphere management plane and additional conditions outside the attacker's control, so the practical risk is concentrated among insiders or attackers who have already compromised privileged credentials rather than the general internet. No public exploit code has been identified at time of analysis, the issue is not listed in CISA KEV, and the vendor has published a security-update advisory (DSA-2026-447) covering this and related PowerMax product vulnerabilities.

Command injection in Dell Unisphere for PowerMax vApp (versions 9.2.4.x) allows a high-privileged remote attacker to execute arbitrary commands, achieving remote code execution. Exploitation requires prior administrative access (PR:H) and high attack complexity (AC:H), limiting risk to insiders or attackers who have already compromised an admin account. No public exploit or active exploitation has been identified, and the CVSS score is 6.6 (moderate).

Command injection and buffer overflow weaknesses in HPE NonStop Java JToolkit can let a locally positioned, already highly privileged attacker corrupt memory or execute unintended commands within the JToolkit process, with impact limited to low confidentiality, integrity, and availability effects (CVSS 3.1 base 3.8, Low). Exploitation is not remote: it requires pre-existing high local privileges (PR:H), local access to the NonStop system (AV:L), a separate user interaction that supplies attacker-influenced input to the vulnerable routine (UI:R), and additional non-deterministic conditions captured by high attack complexity (AC:H). No public exploit code has been identified at time of analysis and there is no confirmed active exploitation (no CISA KEV listing), so all available CVSS signals point to a low real-world priority.

Local privilege escalation in Akamai Guardicore Platform Agent (GPA) on macOS before version 7.4 allows an attacker with administrative privileges to execute arbitrary commands as root by exploiting a command injection in the wapptunneld component. The vulnerability requires local access and high privileges (PR:H), and exploitation necessitates that the 'Log out from your IdP' action is triggered after a malicious network interface name is set. No public exploit code or active exploitation has been identified; this is a genuine but bounded risk limited to unpatched macOS deployments.

Persistent command execution in Citrix UniconOS lets a locally authenticated low-privilege user plant commands that survive normal operation and execute in a privileged context, giving the attacker sustained access to data and code on the endpoint. The flaw affects UniconOS releases before 2607.2 CR and before 2603 LTSR CU2; it is not remotely exploitable and requires an existing foothold or valid low-privilege credentials on the device, with no user interaction needed. No public exploit code has been identified at time of analysis, no confirmed active exploitation (CISA KEV) has been reported, and the vendor has released a patch, making this a moderate-risk, locally-scoped issue rather than an urgent widespread threat.

The `periphery` agent's container-management functionality of Komodo version 2.2.0 contains an OS command injection vulnerability. An authenticated low-privileged user with `Execute`, or `Read` plus `Logs`, permission on a managed server can supply shell metacharacters in a container name processed by the `DestroyContainer` or `SearchContainerLog` operations. The container name is interpolated without escaping into a command executed using `sh -c`, allowing arbitrary commands to execute as root on the managed host. This permits a user without Komodo's `Terminal` permission to obtain unrestricted host-level command execution.

reNgine up to v2.2.0 and current master branch (commit de4199222cca5618f1255930ce78987e47d43a99) is vulnerable to OS Command Injection via the custom_headers scan engine configuration field in web/reNgine/tasks.py (fetch_url and dir_file_fuzz functions).

OS command injection in Fortinet FortiManager (8.0.0 and 7.6.0-7.6.6) and FortiManager Cloud (7.6.2-7.6.6) allows a high-privileged authenticated attacker on an adjacent network to execute arbitrary operating system commands, enabling privilege escalation and potential full compromise of the management appliance. The vulnerability requires an existing administrative account (CVSS PR:H) and network adjacency (AV:A), and no public exploit identified at time of analysis. Despite the high-value nature of the target, these prerequisites substantially narrow realistic risk to post-compromise or malicious insider scenarios, consistent with the moderate-risk assessment.

Authenticated administrators on the local network of NETGEAR CBR40, R7000, R7000P, and RAX40 routers can execute unauthorized commands via insufficient input validation in the NTP time server setting, affecting only router integrity (no confidentiality or availability impact). This medium-severity vulnerability (CVSS 4.0 base 4.3) requires the attacker to already possess administrator credentials and be connected to the router's local network, making it a post-compromise integrity concern rather than a remote attack vector. No public exploit code or active exploitation has been identified, and NETGEAR has released firmware updates to address the issue.

Command injection in Toonflow-app 1.1.8 and earlier allows a local, authenticated low-privileged user to execute arbitrary commands by manipulating the 'path' argument of the Open Folder endpoint (src/routes/setting/fileManagement/openFolder.ts), which is passed without sanitization to child_process.exec. Publicly available exploit code exists, but the attack requires local access and the injected commands run at the privileges of the user already running the application, limiting impact. The vendor-released patch is version 2.0.0; the vendor further states that the affected code paths were removed in the architectural refactor leading to the current 2.0.4 release, so up-to-date installations are not exposed.

Command injection in SAP Yeoman UI (app-studio-toolkit) enables an attacker to execute arbitrary OS commands with the developer's privileges by crafting a malicious workspace settings.json that sets ApplicationWizard.installationLocation to a shell payload. The command fires when the developer triggers a generator installation or uninstallation, which invokes child_process.execSync with the unsanitized value. The assessed CVSS vector (AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H) indicates local exploitation requiring user interaction, with high impact to confidentiality, integrity, and availability on the developer workstation; no confirmed active exploitation or public exploit code is identified in the provided data.

Unauthenticated OS command injection in GPT-SoVITS through 20250606v2pro lets any network-reachable attacker run arbitrary commands as the server user by inserting shell substitutions such as $(...) into the UVR5 WebUI output folder textbox and then triggering an mp3 or m4a export. The root cause (CWE-78) sits in tools/uvr5/vr.py, which hands the user-controlled output path to a shell during export without sanitization, so the impact is full compromise of the account running the WebUI (high confidentiality, integrity and availability loss per both the supplied CVSS 4.0 score of 9.3 and the independent CVSS 3.1 assessment of AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H). Exploitation depends on the UVR5 WebUI being reachable over the network, which is not guaranteed because these Gradio-style instances are commonly bound to localhost or a trusted LAN, and no public exploit code or confirmed in-the-wild activity has been identified from the available intelligence.

Unauthenticated remote command injection in the end-of-life Edimax BR-6208AC router (firmware 1.32) allows an attacker who can reach the device's HTTP web-management interface to execute arbitrary operating-system commands by manipulating the submit-url argument of the /goform/formAccept endpoint. The assessed CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:N, C:H/I:H/A:H) confirms that no credentials and no user interaction are required, and publicly available exploit code exists in the GitHub repository referenced by the disclosure. Edimax was contacted before publication but did not respond and has shipped no patched firmware; the original CVSS 4.0 score of 6.9 understates the confidentiality, integrity and availability impact compared with the independent assessment, and overall exploitability is governed mainly by whether the management interface is reachable from the WAN or only from the local network, making this a moderate real-world priority.

Authenticated administrators of Thelia 2.6.0 and earlier can achieve remote code execution by injecting malicious Smarty expressions into email templates, which execute with web process privileges when previewed or rendered. The vulnerability requires high-privilege back-office access, but successful exploitation enables arbitrary OS command execution and file read/write. No active exploitation has been confirmed and no public exploit code is currently identified.

OS command injection in the Agno agent framework's CodingTools.run_shell helper (versions up to and including 3.0.9) lets an authenticated caller who can influence the tool's `command` argument execute arbitrary shell commands on the host running the agent. The weakness is a bypass of the `restrict_to_base_dir` allowlist and metacharacter block rather than an absence of one: allowlisted interpreters such as python, pip and git can be invoked with inline-code options (for example `python -c`) to run arbitrary code, so the restriction only reduces accidental damage and is explicitly documented as 'not a security sandbox'. Publicly available exploit code exists, but the practical risk is LOW: exploitation requires run_shell to be enabled (it was on by default through 3.0.9 and is off by default from 3.0.10), an agent wired to accept untrusted input, and the affected CVSS 3.1 vector is AV:N/AC:L/PR:L/UI:R/S:U/C:L/I:L/A:L, with the vendor-supplied CVSS 4.0 score of 2.0 consistent with that low assessment. There is no indication of confirmed active exploitation (not in CISA KEV), and a vendor patch is available - upgrade to 3.0.10.

OS command injection in the PAM session-cleanup path of Brocade Fabric OS (before 9.2.2d, and 10.0.0 through 10.0.0a1) lets an authenticated user whose directory- or AAA-supplied username or profile identifier contains shell metacharacters execute arbitrary commands as root when their remote SSH session terminates. Three prerequisites must hold simultaneously: the switch must authenticate users against an external directory or AAA service (LDAP/RADIUS/TACACS+) rather than local accounts only, the attacker must be able to set or influence the username/profile identifier in that directory, and the attacker must be able to establish and then close an authenticated SSH session - so switches using only local authentication are not exposed. No public exploit has been identified at time of analysis; vendor-released patch: 9.2.2d for the 9.x branch, with the corresponding fixed 10.0.x release described in the Broadcom/Brocade advisory at https://support.broadcom.com/external/content/SecurityAdvisories/0/39147.

OS command injection in Brocade Fabric OS account management affects versions before 9.2.2d and 10.0.0 through 10.0.0a1. An authenticated administrator with adjacent management-network access can instigate account deletion, and if the target account has a malformed name containing shell metacharacters-accepted by older Fabric OS builds-the internal cryptographic-key cleanup routine may execute injected commands. This is a medium-severity (CVSS 4.0 base 5.4) issue with low-to-moderate real-world risk because exploitation requires admin privileges, adjacent access, an operator UI action, and the non-default precondition of a pre-existing malformed account name; no public exploit has been identified at time of analysis.

OS command injection in the web-based management interface of Cisco License On-Prem (formerly Cisco Smart Software Manager On-Prem) allows an authenticated remote attacker holding system administrator credentials to submit a crafted configuration update that is executed as root on the underlying operating system. Scored CVSS 4.9 (AV:N/AC:L/PR:H/UI:N/S:U/C:N/I:N/A:H), the flaw is heavily constrained by its PR:H authentication prerequisite: because the attacker must already be a full administrator, the marginal capability gained is limited to shutting down or disrupting the host (availability impact only), which the assessed vector reflects with C:N/I:N/A:H. No public exploit code was identified at time of analysis and the issue is not in CISA KEV; no EPSS score was supplied in the source data.

OS command injection in the web-based management API of Cisco Application Policy Infrastructure Controller (APIC) allows an attacker who already holds valid administrative credentials to run arbitrary commands as root on the underlying controller operating system. Exploitation is remote and requires no user interaction, but the PR:H rating in the CVSS vector confirms the high privilege bar: only authenticated admins can reach the vulnerable code path, so realistic exposure is limited to insider abuse, stolen or shared admin accounts, or APIC management interfaces reachable from untrusted networks. No public exploit code has been identified at time of analysis and there is no CISA KEV entry; Cisco's base score of 6.5 understates the assessed impact, since the independent re-scoring (CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H) also credits full loss of availability on the controller.

Unauthenticated command injection in the Fanvil X7A IP phone's web management portal (firmware 2.6.0.1182) lets any attacker who can reach the diagnostic ping tool execute arbitrary commands with the privileges of the underlying Android operating system, resulting in total device compromise (CVSS 10.0, scope-changed, S:C). No credentials or user interaction are required (PR:N/UI:N), publicly available exploit code exists, and CISA's SSVC assessment rates the flaw fully automatable with total technical impact, so any deployment with the portal reachable from untrusted or flat networks should be assumed exploitable at scale. In correctly segmented environments the real-world attacker population is limited to hosts with management-VLAN or LAN reachability to the phone's portal.

Unauthenticated command injection in the Fanvil x7a IP phone's web management portal (firmware 2.6.0.1182) allows any attacker who can reach the portal's diagnostic ping tool to inject shell metacharacters and execute arbitrary commands on the underlying Android operating system, compromising confidentiality, integrity and availability equally (CVSS 9.8, AV:N/AC:L/PR:N/UI:N). The practical gating factor is exposure: the portal is normally reachable only from a LAN or management VLAN, so risk falls sharply wherever it is not reachable from untrusted networks. No public exploit code or confirmed in-the-wild activity has been identified at time of analysis - CISA's SSVC rates Exploitation as 'none' while classifying the flaw as automatable with total technical impact, meaning exploitation could spread quickly once a proof-of-concept surfaces.

Command injection in the HPE Aruba ClearPass Policy Manager OnGuard agent allows an attacker holding a valid authenticated session within a ClearPass-managed environment to inject and execute arbitrary operating-system commands, which per the vulnerability description run with elevated privileges on the affected Windows endpoint. Only deployments that have installed the OnGuard posture/health agent on Windows hosts are exposed; non-Windows endpoints and environments not using OnGuard are out of scope. HPE rates this medium severity (CVSS 6.5), and our independent assessment keeps the severity in that band while noting that, once the authentication prerequisite is satisfied, the impact on the endpoint is high because the injected commands inherit elevated rights (assessed confidentiality, integrity and availability all High). No public exploit code has been identified and CISA KEV does not list this issue, so there is no confirmation of active exploitation at time of analysis.