A stack-based buffer overflow vulnerability exists in the dataflowstatus function of the /cgi-bin/mainfunction.cgi component in DrayTek Vigor2960 firmware versions up to 1.5.1.6. This flaw can be exploited remotely by manipulating the argument option, potentially leading to complete compromise. An exploit has been published. Upgrading to version 1.5.1.7 mitigates this issue.
A critical stack-based buffer overflow vulnerability exists in the doWebBackup function of the /cgi-bin/mainfunction.cgi component in DrayTek Vigor2960 firmware versions up to 1.5.1.6. This vulnerability can be exploited remotely by manipulating the path argument, potentially leading to full system compromise. Public exploit code has been disclosed. Upgrading to version 1.5.1.7 resolves the issue.
A critical stack-based buffer overflow vulnerability exists in the json_load_lock function of the /cgi-bin/mainfunction.cgi component (Jsonstatus) in DrayTek Vigor2960 firmware up to version 1.5.1.6. This vulnerability can be exploited remotely by manipulating the 'usefile' argument. Public exploit code is available. Upgrading to version 1.5.1.7 resolves the issue.
CVE-2026-132560 is an information disclosure vulnerability in MemTensor memmy-agent versions 1.1.0 and 1.1.1. The flaw exists in the /api/v1/memory/logs endpoint of the Memory HTTP Service component. Exploitation requires local access and can lead to unauthorized information disclosure. A patch identified by commit 421ce6cb33861255d34c3887d39aced009b3ce9b is available to address this issue. The vulnerability has a medium severity score of 4.8 and public exploit code has been published.
Red Hat has released a critical security update for the java-25-openjdk packages in Red Hat Enterprise Linux 9. This update addresses multiple vulnerabilities in the OpenJDK 25 Java Runtime Environment and Software Development Kit, including enhancements and fixes related to Kerberos tickets, TLS connections, StringBuilder checks, JFR events, loop handling, vector masks, integer compression, JMX connections, and Base64 optimization. Additionally, a bug fix disables post-quantum hybrid key exchange in TLS 1.3 when running in FIPS mode to prevent negotiation failures.
CVE-2026-91139 is a cryptographic vulnerability in TP-Link Systems Inc. Deco BE65 v3 firmware. The local web-management authentication flow uses an insufficient 512-bit RSA key, which weakens the cryptographic protection. This weakness could allow attackers to compromise the authentication traffic. The issue affects version 0 of the firmware and is fixed in version 1.0.11 Build 26092212 Rel. 18351. The CVSS 4.0 score is 6.1, indicating a medium severity level.
CVE-2026-91138 is a vulnerability in TP-Link Systems Inc. Deco BE65 v3 firmware where a hard-coded password hash for the root account is present in /etc/shadow. An attacker with unprivileged shell access may extract the plaintext password from this hash and escalate privileges to root. This flaw allows unauthorized root-level access to the device. A fix is available in firmware version 1.0.11 Build 26092212 Rel. 18351.
Vtiger CRM versions up to 8.4.0 contain a blind SQL injection vulnerability in the ReportRun::generateAdvFilterSql() function. This flaw allows authenticated users with low privileges to inject SQL code via the 'bw' comparator in the advanced_filter parameter. Exploitation involves sending a crafted parameter to the Reports getRecordsCount action to break out of SQL literals and extract database information. The vulnerability has a high severity score of 7.1 (CVSS 4.0).
CVE-2026-132750 is a high-severity vulnerability in Meilisearch before versions 1.55.1 and 1.56.1. It involves a quadratic memory consumption issue in the filter-parser component, allowing attackers with a search API key to exhaust server memory by sending large filters, such as a _geoPolygon with thousands of coordinates. This can crash the server and take all indexes offline.
A NULL pointer dereference vulnerability exists in TechSmith mp4v2 versions up to 2.1.0 in the MP4RtpHintTrack::GetPayload() function. This flaw occurs when an unchecked rtpmap value is passed to strchr(), allowing attackers to craft MP4 files with corrupted RTP hint tracks that can crash mp4info or other applications using MP4GetHintTrackRtpPayload().
Geany versions up to 2.1.0 contain an OS command injection vulnerability. This flaw allows attackers to execute arbitrary shell commands by crafting filenames with special shell metacharacters such as backticks or $(...). When a user runs Compile, Build, or Execute commands on such files, the embedded commands execute with the user's privileges.
CVE-2026-71534 is a vulnerability in mareforma versions 0.3.0 through 0.3.7 where the restore logic improperly derives the transparency_logged flag from an unsigned field without verifying claim-bound entries. This allows an attacker to manipulate claims.toml to falsely mark claims as eligible for REPLICATED convergence. Additionally, in non-Rekor graphs, the logic resets signed claims from transparency_logged 1 to 0 after restore, preventing convergence permanently. The issue is fixed in version 0.3.8.
Dokploy versions prior to 0.31.1 contain a critical OS command injection vulnerability in the Nixpacks publishDirectory build setting. This setting improperly handles shell metacharacters, which are interpolated into a docker cp command executed via child_process.exec. An authenticated user with application access can exploit this to execute arbitrary commands on the Dokploy host, potentially leading to full host compromise. The vulnerability is fixed in version 0.31.1.
Dokploy versions prior to 0.31.1 contain a missing authorization vulnerability in network-related procedures. These procedures require only authentication but not the Docker permission, allowing standard members to perform sensitive network operations such as listing, inspecting, creating, importing, recreating, resynchronizing, and deleting Docker networks. This can lead to disruption of service connectivity and unauthorized network configuration changes. The vulnerability is fixed in version 0.31.1.
Dokploy versions prior to 0.31.1 contain a command injection vulnerability in the mount.deleteFileMount function. An authenticated user with file-mount create and delete permissions can execute arbitrary commands on the host by deleting a local file mount. This occurs because the user-controlled filePath is interpolated unquoted into an rm -rf command. The issue is fixed in version 0.31.1.
Dokploy versions prior to 0.31.1 contain a command injection vulnerability (CWE-77) in the handling of user-controlled schedule appName values. Authenticated users able to create or update schedules can inject OS commands executed with Dokploy process privileges, potentially including root on typical installations. This vulnerability is fixed in version 0.31.1.