Open source vulnerabilities

GitLab has remediated an issue in GitLab CE/EE affecting all versions from 18.7 before 18.11.12, 19.0 before 19.0.9, 19.1 before 19.1.8, 19.2 before 19.2.6, and 19.3 before 19.3.2 that, under certain conditions, an unauthenticated user could have read arbitrary files from the GitLab server due to improper path confinement and missing authentication enforcement in the repository commits API.

WordPress 6.9.x before 6.9.5 and 7.0.x before 7.0.2 is affected by a REST API batch endpoint route confusion issue which, combined with the author__not_in WP_Query SQL Injection (CVE-2026-60137), could allow an attacker to perform SQL Injection and achieve Remote Code Execution.

WordPress 6.8.x before 6.8.6, 6.9.x before 6.9.5, and 7.0.x before 7.0.2 does not properly sanitise the author__not_in parameter of WP_Query, which could allow SQL Injection when a plugin or theme passes untrusted input to the parameter.

An issue was discovered in router/upnp/src/ssdp.c in DD-WRT before 45724. An unsafe strcpy in the UPnP handling functionality allows an unauthenticated remote attacker to send a request that would overflow an internal fixed buffer. Exploitation requires the DD-WRT user to enable UPnP (which is off by default, and only listens on internal interfaces by default). This occurs in ssdp_msearch (reachable by an M-SEARCH request).

authentik is an open-source identity provider. Prior to versions 2025.12.5 and 2026.2.3, due to the implementation of stages in the SFE (Simple Flow Executor) in order to make the interface more compatible with legacy browsers, it was possible to use an XSS exploit in the AutosubmitStage. This issue has been patched in versions 2025.12.5 and 2026.2.3.

Roundcube Webmail 1.6.x before 1.6.16 and 1.7.x before 1.7.1 has Pre-authentication SQL injection in the virtuser_query plugin via a preg_replace() backslash escape bypass.

A flaw has been found in opensourcepos Open Source Point of Sale up to 3.4.2. Impacted is the function Login of the file app/Models/Employee.php of the component Employee Login. This manipulation causes use of weak hash. Remote exploitation of the attack is possible. The attack is considered to have high complexity. The exploitability is considered difficult. The actual existence of this vulnerability is currently in question. The vendor explains: "[T]he code is still there to allow the upgrade path to work. The default password is initially seeded with the old hash function, but then migrated to a newer one after login. [T]he hash version check might be cleaned up in the future. Currently it's not actively in use as any password change will use a newer hash function."

A vulnerability was detected in opensourcepos Open Source Point of Sale up to 3.4.2. This issue affects the function getPicThumb of the file app/Controllers/Items.php. The manipulation of the argument pic_filename results in path traversal. The attack may be launched remotely. The patch is identified as def0c27a0e252668df8d942fc31e16d1edfd7323. A patch should be applied to remediate this issue. The vendor was contacted early about this disclosure.

MinIO is a high-performance object storage system. From RELEASE.2022-07-24T01-54-52Z to before RELEASE.2026-04-14T21-32-45Z, A path traversal vulnerability in MinIO's ReadMultiple internode storage-REST endpoint allows a caller holding the cluster root JWT to read files from outside the configured drive roots, bounded only by the MinIO process UID. The attacker sends POST minio/storage/{drivePath}/v63/rmpl with a msgpack-encoded body carrying ../ sequences in the Bucket field. The server opens the resulting path via os.OpenFile with O_RDONLY|O_NOATIME and returns its contents in the msgpack response stream. This vulnerability is fixed in RELEASE.2026-04-14T21-32-45Z.

OpenPrinting CUPS is an open source printing system for Linux and other Unix-like operating systems. Prior to 2.4.17, a network-adjacent attacker can send a crafted SNMP response to the CUPS SNMP backend that causes an out-of-bounds read of up to 176 bytes past a stack buffer. The leaked memory is converted from UTF-16 to UTF-8 and stored as printer supply description strings, which are subsequently visible to authenticated users via IPP Get-Printer-Attributes responses and the CUPS web interface. This vulnerability is fixed in 2.4.17.

MinIO is a high-performance object storage system. Starting in RELEASE.2023-05-18T00-05-36Z and prior to RELEASE.2026-04-11T03-20-12Z, an authentication bypass vulnerability in MinIO's `STREAMING-UNSIGNED-PAYLOAD-TRAILER` code path allows any user who knows a valid access key to write arbitrary objects to any bucket without knowing the secret key or providing a valid cryptographic signature. Any MinIO deployment is impacted. The attack requires only a valid access key (the well-known default `minioadmin`, or any key with WRITE permission on a bucket) and a target bucket name. `PutObjectHandler` and `PutObjectPartHandler` call `newUnsignedV4ChunkedReader` with a signature verification gate based solely on the presence of the `Authorization` header. Meanwhile, `isPutActionAllowed` extracts credentials from either the `Authorization` header or the `X-Amz-Credential` query parameter, and trusts whichever it finds. An attacker omits the `Authorization` header and supplies credentials exclusively via the query string. The signature gate evaluates to `false`, `doesSignatureMatch` is never called, and the request proceeds with the permissions of the impersonated access key. This affects `PutObjectHandler` (standard and tables/warehouse bucket paths) and `PutObjectPartHandler` (multipart uploads). Users of the open-source `minio/minio` project should upgrade to MinIO AIStor `RELEASE.2026-04-11T03-20-12Z` or later. If upgrading is not immediately possible, block unsigned-trailer requests at the load balancer. Reject any request containing `X-Amz-Content-Sha256: STREAMING-UNSIGNED-PAYLOAD-TRAILER` at the reverse proxy or WAF layer. Clients can use `STREAMING-AWS4-HMAC-SHA256-PAYLOAD-TRAILER` (the signed variant) instead. Alternatively, restrict WRITE permissions. Limit `s3:PutObject` grants to trusted principals. While this reduces the attack surface, it does not eliminate the vulnerability since any user with WRITE permission can exploit it with only their access key.

MinIO is a high-performance object storage system. Starting in RELEASE.2023-05-18T00-05-36Z and prior to RELEASE.2026-04-11T03-20-12Z, an authentication bypass vulnerability in MinIO's Snowball auto-extract handler (`PutObjectExtractHandler`) allows any user who knows a valid access key to write arbitrary objects to any bucket without knowing the secret key or providing a valid cryptographic signature. Any MinIO deployment is impacted. The attack requires only a valid access key (the well-known default `minioadmin`, or any key with WRITE permission on a bucket) and a target bucket name. When `authTypeStreamingUnsignedTrailer` support was added, the new auth type was handled in `PutObjectHandler` and `PutObjectPartHandler` but was never added to `PutObjectExtractHandler`. The snowball auto-extract handler's `switch rAuthType` block has no case for `authTypeStreamingUnsignedTrailer`, so execution falls through with zero signature verification. The `isPutActionAllowed` call before the switch extracts the access key and checks IAM permissions, but does not verify the cryptographic signature. An attacker sends a PUT request with `X-Amz-Content-Sha256: STREAMING-UNSIGNED-PAYLOAD-TRAILER`, `X-Amz-Meta-Snowball-Auto-Extract: true`, and an `Authorization` header containing a valid access key with a completely fabricated signature. The request is accepted and the tar payload is extracted into the bucket. Users of the open-source minio/minio project should upgrade to MinIO AIStor RELEASE.2026-04-11T03-20-12Z or later. If upgrading is not immediately possible, block unsigned-trailer requests at the load balancer. Reject any request containing X-Amz-Content-Sha256: STREAMING-UNSIGNED-PAYLOAD-TRAILER at the reverse proxy or WAF layer. Clients can use STREAMING-AWS4-HMAC-SHA256-PAYLOAD-TRAILER (the signed variant) instead. Alternatively, restrict WRITE permissions. Limit s3:PutObject grants to trusted principals. While this reduces the attack surface, it does not eliminate the vulnerability since any user with WRITE permission can exploit it with only their access key.

Open Source Point of Sale is a web based point-of-sale application written in PHP using CodeIgniter framework. Prior to 3.4.3, a Stored Cross-Site Scripting (XSS) vulnerability exists in the Daily Sales management table. The customer_name column is configured with escape: false in the bootstrap-table column configuration, causing customer names to be rendered as raw HTML. An attacker with customer management permissions can inject arbitrary JavaScript into a customer's first_name or last_name field, which executes in the browser of any user viewing the Daily Sales page. This vulnerability is fixed in 3.4.3.

OpenPrinting CUPS is an open source printing system for Linux and other Unix-like operating systems. In versions 2.4.16 and prior, CUPS daemon (cupsd) contains an authorization bypass vulnerability due to case-insensitive username comparison during authorization checks. The vulnerability allows an unprivileged user to gain unauthorized access to restricted operations by using a user with a username that differs only in case from an authorized user. At time of publication, there are no publicly available patches.

Open Source Point of Sale (opensourcepos) is a web based point of sale application written in PHP using CodeIgniter framework. Prior to version 3.4.2, an Insecure Direct Object Reference (IDOR) vulnerability allows an authenticated low-privileged user to access the password change functionality of other users, including administrators, by manipulating the `employee_id` parameter. The application does not verify object ownership or enforce authorization checks. Version 3.4.2 adds object-level authorization checks to validate that the current user owns the employee_id being accessed.

The Delete function fails to properly validate offsets when processing malformed JSON input. This can lead to a negative slice index and a runtime panic, allowing a denial of service attack.

Mesa is an open-source Python library for agent-based modeling, simulating complex systems and exploring emergent behaviors. In version 3.5.0 and prior, checking out of untrusted code in benchmarks.yml workflow may lead to code execution in privileged runner. This issue has been patched via commit c35b8cd.

OpenSourcePOS 3.4.1 has a second order SQL Injection vulnerability in the handling of the currency_symbol configuration field. Although the input is initially stored without immediate execution, it is later concatenated into a dynamically constructed SQL query without proper sanitization or parameter binding. This allows an attacker with access to modify the currency_symbol value to inject arbitrary SQL expressions, which are executed when the affected query is subsequently processed.

An issue in Visual Studio Code Extensions Live Server v5.7.9 allows attackers to exfiltrate files via user interaction with a crafted HTML page.

An issue in Visual Studio Code Extensions Markdown Preview Enhanced v0.8.18 allows attackers to execute arbitrary code via uploading a crafted .Md file.

An issue in the code-runner.executorMap setting of Visual Studio Code Extensions Code Runner v0.12.2 allows attackers to execute arbitrary code when opening a crafted workspace.

Gogs is an open source self-hosted Git service. In version 0.13.3 and prior, a path traversal vulnerability exists in the updateWikiPage function of Gogs. The vulnerability allows an authenticated user with write access to a repository's wiki to delete arbitrary files on the server by manipulating the old_title parameter in the wiki editing form. This issue has been patched in versions 0.13.4 and 0.14.0+dev.

Gogs is an open source self-hosted Git service. In version 0.13.3 and prior, there is an arbitrary file read/write via path traversal in Git hook editing. This issue has been patched in versions 0.13.4 and 0.14.0+dev.

Gogs is an open source self-hosted Git service. In version 0.13.3 and prior, the endpoint "PUT /repos/:owner/:repo/contents/*" does not require write permissions and allows access with read permission only via repoAssignment(). After passing the permission check, PutContents() invokes UpdateRepoFile(), which results in commit creation and the execution of git push. As a result, a token with read-only permission can be used to modify repository contents. This issue has been patched in versions 0.13.4 and 0.14.0+dev.

Langflow exec_globals Inclusion of Functionality from Untrusted Control Sphere Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Langflow. Authentication is not required to exploit this vulnerability. The specific flaw exists within the handling of the exec_globals parameter provided to the validate endpoint. The issue results from the inclusion of a resource from an untrusted control sphere. An attacker can leverage this vulnerability to execute code in the context of root. Was ZDI-CAN-27325.

OpenProject is an open-source, web-based project management software. Versions 16.3.0 through 16.6.4 are affected by a stored cross-site scripting vulnerability in the Roadmap view. OpenProject’s roadmap view renders the “Related work packages” list for each version. When a version contains work packages from a different project (e.g., a subproject), the helper link_to_work_package prepends package.project.to_s to the link and returns the entire string with .html_safe. Because project names are user-controlled and no escaping happens before calling html_safe, any HTML placed in a subproject name is injected verbatim into the page. The underlying issue is mitigated in versions 16.6.5 and 17.0.0 by setting a `X-Content-Type-Options: nosniff` header, which was in place until a refactoring move to Rails standard content-security policy, which did not properly apply this header in the new configuration since OpenProject 16.3.0. Those who cannot upgrade their installations should ensure that they add a X-Content-Type-Options: nosniff header in their proxying web application server.

In mcp-server-git versions prior to 2025.12.17, when the server is started with the --repository flag to restrict operations to a specific repository path, it did not validate that repo_path arguments in subsequent tool calls were actually within that configured path. This could allow tool calls to operate on other repositories accessible to the server process. The fix adds path validation that resolves both the configured repository and the requested path (following symlinks) and verifies the requested path is within the allowed repository before executing any git operations. Users are advised to upgrade to 2025.12.17 upon release to remediate this issue.

In mcp-server-git versions prior to 2025.12.17, the git_diff and git_checkout functions passed user-controlled arguments directly to git CLI commands without sanitization. Flag-like values (e.g., `--output=/path/to/file` for `git_diff`) would be interpreted as command-line options rather than git refs, enabling arbitrary file overwrites. The fix adds validation that rejects arguments starting with - and verifies the argument resolves to a valid git ref via rev_parse before execution. Users are advised to update to 2025.12.17 resolve this issue when it is released.

Model Context Protocol Servers is a collection of reference implementations for the model context protocol (MCP). In mcp-server-git versions prior to 2025.9.25, the git_init tool accepted arbitrary filesystem paths and created Git repositories without validating the target location. Unlike other tools which required an existing repository, git_init could operate on any directory accessible to the server process, making those directories eligible for subsequent git operations. The tool was removed entirely, as the server is intended to operate on existing repositories only. Users are advised to upgrade to 2025.9.25 or newer to remediate this issue.

Improper Neutralization of Escape, Meta, or Control Sequences vulnerability in Apache Tomcat. Tomcat did not escape ANSI escape sequences in log messages. If Tomcat was running in a console on a Windows operating system, and the console supported ANSI escape sequences, it was possible for an attacker to use a specially crafted URL to inject ANSI escape sequences to manipulate the console and the clipboard and attempt to trick an administrator into running an attacker controlled command. While no attack vector was found, it may have been possible to mount this attack on other operating systems. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.10, from 10.1.0-M1 through 10.1.44, from 9.0.40 through 9.0.108. The following versions were EOL at the time the CVE was created but are known to be affected: 8.5.60 though 8.5.100. Other, older, EOL versions may also be affected. Users are recommended to upgrade to version 11.0.11 or later, 10.1.45 or later or 9.0.109 or later, which fix the issue.

Relative Path Traversal vulnerability in Apache Tomcat. The fix for bug 60013 introduced a regression where the rewritten URL was normalized before it was decoded. This introduced the possibility that, for rewrite rules that rewrite query parameters to the URL, an attacker could manipulate the request URI to bypass security constraints including the protection for /WEB-INF/ and /META-INF/. If PUT requests were also enabled then malicious files could be uploaded leading to remote code execution. PUT requests are normally limited to trusted users and it is considered unlikely that PUT requests would be enabled in conjunction with a rewrite that manipulated the URI. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.10, from 10.1.0-M1 through 10.1.44, from 9.0.0.M11 through 9.0.108. The following versions were EOL at the time the CVE was created but are known to be affected: 8.5.6 though 8.5.100. Other, older, EOL versions may also be affected. Users are recommended to upgrade to version 11.0.11 or later, 10.1.45 or later or 9.0.109 or later, which fix the issue.

MinIO is a high-performance object storage system. In all versions prior to RELEASE.2025-10-15T17-29-55Z, a privilege escalation vulnerability allows service accounts and STS (Security Token Service) accounts with restricted session policies to bypass their inline policy restrictions when performing operations on their own account, specifically when creating new service accounts for the same user. The vulnerability exists in the IAM policy validation logic where the code incorrectly relied on the DenyOnly argument when validating session policies for restricted accounts. When a session policy is present, the system should validate that the action is allowed by the session policy, not just that it is not denied. An attacker with valid credentials for a restricted service or STS account can create a new service account for itself without policy restrictions, resulting in a new service account with full parent privileges instead of being restricted by the inline policy. This allows the attacker to access buckets and objects beyond their intended restrictions and modify, delete, or create objects outside their authorized scope. The vulnerability is fixed in version RELEASE.2025-10-15T17-29-55Z.

Rob--W cors-anywhere instances configured as an open proxy allow unauthenticated external users to induce the server to make HTTP requests to arbitrary targets (SSRF). Because the proxy forwards requests and headers, an attacker can reach internal-only endpoints and link-local metadata services, retrieve instance role credentials or other sensitive metadata, and interact with internal APIs and services that are not intended to be internet-facing. The vulnerability is exploitable by sending crafted requests to the proxy with the target resource encoded in the URL; many cors-anywhere deployments forward arbitrary methods and headers (including PUT), which can permit exploitation of IMDSv2 workflows as well as access to internal management APIs. Successful exploitation can result in theft of cloud credentials, unauthorized access to internal services, remote code execution or privilege escalation (depending on reachable backends), data exfiltration, and full compromise of cloud resources. Mitigation includes: restricting the proxy to trusted origins or authentication, whitelisting allowed target hosts, preventing access to link-local and internal IP ranges, removing support for unsafe HTTP methods/headers, enabling cloud provider mitigations, and deploying network-level protections.

NVIDIA Triton Inference Server for Windows and Linux contains a vulnerability in the Python backend, where an attacker could cause an out-of-bounds read by sending a request. A successful exploit of this vulnerability might lead to information disclosure.

NVIDIA Triton Inference Server for Windows and Linux contains a vulnerability in the Python backend, where an attacker could cause the shared memory limit to be exceeded by sending a very large request. A successful exploit of this vulnerability might lead to information disclosure.

NVIDIA Triton Inference Server for Windows and Linux contains a vulnerability in the Python backend, where an attacker could cause an out-of-bounds write by sending a request. A successful exploit of this vulnerability might lead to remote code execution, denial of service, data tampering, or information disclosure.

mcp-remote is exposed to OS command injection when connecting to untrusted MCP servers due to crafted input from the authorization_endpoint response URL

Remote Code Execution in API component in Ivanti Endpoint Manager Mobile 12.5.0.0 and prior on unspecified platforms allows authenticated attackers to execute arbitrary code via crafted API requests.

An authentication bypass in the API component of Ivanti Endpoint Manager Mobile 12.5.0.0 and prior allows attackers to access protected resources without proper credentials via the API.

Craft is a flexible, user-friendly CMS for creating custom digital experiences on the web and beyond. Starting from version 3.0.0-RC1 to before 3.9.15, 4.0.0-RC1 to before 4.14.15, and 5.0.0-RC1 to before 5.6.17, Craft is vulnerable to remote code execution. This is a high-impact, low-complexity attack vector. This issue has been patched in versions 3.9.15, 4.14.15, and 5.6.17, and is an additional fix for CVE-2023-41892.

Path Equivalence: 'file.Name' (Internal Dot) leading to Remote Code Execution and/or Information disclosure and/or malicious content added to uploaded files via write enabled Default Servlet in Apache Tomcat. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.2, from 10.1.0-M1 through 10.1.34, from 9.0.0.M1 through 9.0.98. The following versions were EOL at the time the CVE was created but are known to be affected: 8.5.0 though 8.5.100. Other, older, EOL versions may also be affected. If all of the following were true, a malicious user was able to view security sensitive files and/or inject content into those files: - writes enabled for the default servlet (disabled by default) - support for partial PUT (enabled by default) - a target URL for security sensitive uploads that was a sub-directory of a target URL for public uploads - attacker knowledge of the names of security sensitive files being uploaded - the security sensitive files also being uploaded via partial PUT If all of the following were true, a malicious user was able to perform remote code execution: - writes enabled for the default servlet (disabled by default) - support for partial PUT (enabled by default) - application was using Tomcat's file based session persistence with the default storage location - application included a library that may be leveraged in a deserialization attack Users are recommended to upgrade to version 11.0.3, 10.1.35 or 9.0.99, which fixes the issue.

/vendor/htmlawed/htmlawed/htmLawedTest.php in the htmlawed module for GLPI through 10.0.2 allows PHP code injection.

The Samba vfs_fruit module uses extended file attributes (EA, xattr) to provide "...enhanced compatibility with Apple SMB clients and interoperability with a Netatalk 3 AFP fileserver." Samba versions prior to 4.13.17, 4.14.12 and 4.15.5 with vfs_fruit configured allow out-of-bounds heap read and write via specially crafted extended file attributes. A remote attacker with write access to extended file attributes can execute arbitrary code with the privileges of smbd, typically root.

Apache Log4j2 2.0-beta9 through 2.15.0 (excluding security releases 2.12.2, 2.12.3, and 2.3.1) JNDI features used in configuration, log messages, and parameters do not protect against attacker controlled LDAP and other JNDI related endpoints. An attacker who can control log messages or log message parameters can execute arbitrary code loaded from LDAP servers when message lookup substitution is enabled. From log4j 2.15.0, this behavior has been disabled by default. From version 2.16.0 (along with 2.12.2, 2.12.3, and 2.3.1), this functionality has been completely removed. Note that this vulnerability is specific to log4j-core and does not affect log4net, log4cxx, or other Apache Logging Services projects.

XStream is a simple library to serialize objects to XML and back again. In affected versions this vulnerability may allow a remote attacker has sufficient rights to execute commands of the host only by manipulating the processed input stream. No user is affected, who followed the recommendation to setup XStream's security framework with a whitelist limited to the minimal required types. XStream 1.4.18 uses no longer a blacklist by default, since it cannot be secured for general purpose.

A change introduced in Apache Flink 1.11.0 (and released in 1.11.1 and 1.11.2 as well) allows attackers to read any file on the local filesystem of the JobManager through the REST interface of the JobManager process. Access is restricted to files accessible by the JobManager process. All users should upgrade to Flink 1.11.3 or 1.12.0 if their Flink instance(s) are exposed. The issue was fixed in commit b561010b0ee741543c3953306037f00d7a9f0801 from apache/flink:master.

DCTStream::getChars in DCTStream.cc in Poppler 20.12.1 has a heap-based buffer overflow via a crafted PDF document. NOTE: later reports indicate that this only affects builds from Poppler git clones in late December 2020, not the 20.12.1 release. In this situation, it should NOT be considered a Poppler vulnerability. However, several third-party Open Source projects directly rely on Poppler git clones made at arbitrary times, and therefore the CVE remains useful to users of those projects

Archive_Tar through 1.4.10 has :// filename sanitization only to address phar attacks, and thus any other stream-wrapper attack (such as file:// to overwrite files) can still succeed.

XML-RPC request are vulnerable to unsafe deserialization and Cross-Site Scripting issues in Apache OFBiz 17.12.03

A remote code execution vulnerability in MobileIron Core & Connector versions 10.3.0.3 and earlier, 10.4.0.0, 10.4.0.1, 10.4.0.2, 10.4.0.3, 10.5.1.0, 10.5.2.0 and 10.6.0.0; and Sentry versions 9.7.2 and earlier, and 9.8.0; and Monitor and Reporting Database (RDB) version 2.0.0.1 and earlier that allows remote attackers to execute arbitrary code via unspecified vectors.