Graphiti is a framework for building and querying temporal context graphs for AI agents. Graphiti versions before 0.28.2 contained a Cypher injection vulnerability in shared search-filter construction for non-Kuzu backends. Attacker-controlled label values supplied through SearchFilters.node_labels were concatenated directly into Cypher label expressions without validation. In MCP deployments, this was exploitable not only through direct untrusted access to the Graphiti MCP server, but also through prompt injection against an LLM client that could be induced to call search_nodes with attacker-controlled entity_types values. The MCP server mapped entity_types to SearchFilters.node_labels, which then reached the vulnerable Cypher construction path. Affected backends included Neo4j, FalkorDB, and Neptune. Kuzu was not affected by the label-injection issue because it used parameterized label handling rather than string-interpolated Cypher labels. This issue was mitigated in 0.28.2.

getzep

Dataease is an open source data visualization analysis tool. Prior to 2.10.20, By controlling the IniFile parameter, an attacker can force the JDBC driver to load an attacker-controlled configuration file. This configuration file can inject dangerous JDBC properties, leading to remote code execution. The Redshift JDBC driver execution flow reaches a method named getJdbcIniFile. The getJdbcIniFile method implements an aggressive automatic configuration file discovery mechanism. If not explicitly restricted, it searches for a file named rsjdbc.ini. In a JDBC URL context, users can explicitly specify the configuration file via URL parameters, which allows arbitrary files on the server to be loaded as JDBC configuration files. Within the Redshift JDBC driver properties, the parameter IniFile is explicitly supported and used to load an external configuration file. This vulnerability is fixed in 2.10.20.

dataease

Hyperterse is a tool-first MCP framework for building AI-ready backend surfaces from declarative config. Prior to v2.2.0, the search tool allows LLMs to search for tools using natural language. While returning results, Hyperterse also returned the raw SQL queries, exposing statements which were supposed to be executed under the hood, and protected from being displayed publicly. This issue has been fixed as of v2.2.0.

hyperterse

Inout EasyRooms Ultimate Edition v1.0 contains an SQL injection vulnerability that allows unauthenticated attackers to manipulate database queries by injecting SQL code through the property1 parameter. Attackers can send POST requests to the search/searchdetailed endpoint with malicious SQL payloads to extract sensitive data or modify database contents.

inoutscripts

Inout EasyRooms Ultimate Edition v1.0 contains an SQL injection vulnerability that allows unauthenticated attackers to manipulate database queries by injecting SQL code through the numguest parameter. Attackers can send POST requests to the search/searchdetailed endpoint with malicious SQL payloads to bypass authentication, extract sensitive data, or modify database contents.

inoutscripts

Inout EasyRooms Ultimate Edition v1.0 contains an SQL injection vulnerability that allows unauthenticated attackers to manipulate database queries by injecting SQL code through the location parameter. Attackers can send POST requests to the search/searchdetailed endpoint with malicious SQL payloads in the location field to extract sensitive data or modify database contents.

inoutscripts

Inout EasyRooms Ultimate Edition v1.0 contains an SQL injection vulnerability that allows unauthenticated attackers to manipulate database queries by injecting SQL code through the guests parameter. Attackers can send POST requests to the search/rentals endpoint with malicious SQL payloads to bypass authentication, extract sensitive data, or modify database contents.

inoutscripts

iScripts ReserveLogic contains an SQL injection vulnerability that allows unauthenticated attackers to manipulate database queries by injecting SQL code through the jqSearchDestination parameter. Attackers can send POST requests to the search endpoint with crafted SQL payloads to extract sensitive database information.

StudioCMS is a server-side-rendered, Astro native, headless content management system. Prior to 0.4.3, the POST /studiocms_api/dashboard/create-reset-link endpoint allows any authenticated user with admin privileges to generate a password reset token for any other user, including the owner account. The handler verifies that the caller is an admin but does not enforce role hierarchy, nor does it validate that the target userId matches the caller's identity. Combined with the POST /studiocms_api/dashboard/reset-password endpoint, this allows a complete account takeover of the highest-privileged account in the system. This vulnerability is fixed in 0.4.3.

studiocms

xygeni-action is the GitHub Action for Xygeni Scanner. On March 3, 2026, an attacker with access to compromised credentials created a series of pull requests (#46, #47, #48) injecting obfuscated shell code into action.yml. The PRs were blocked by branch protection rules and never merged into the main branch. However, the attacker used the compromised GitHub App credentials to move the mutable v5 tag to point at the malicious commit (4bf1d4e19ad81a3e8d4063755ae0f482dd3baf12) from one of the unmerged PRs. This commit remained in the repository's git object store, and any workflow referencing @v5 would fetch and execute it. This is a supply chain compromise via tag poisoning. Any GitHub Actions workflow referencing xygeni/xygeni-action@v5 during the affected window (approximately March 3–10, 2026) executed a C2 implant that granted the attacker arbitrary command execution on the CI runner for up to 180 seconds per workflow run.

xygeni

WeGIA is a web manager for charitable institutions. In 3.6.5, The patched loadBackupDB() extracts tar.gz archives to a temporary directory using PHP's PharData class, then uses glob() and file_get_contents() to read SQL files from the extracted contents. Neither the extraction nor the file reading validates whether archive members are symbolic links. This vulnerability is fixed in 3.6.6.

wegia

FileThingie 2.5.7 contains an arbitrary file upload vulnerability that allows attackers to upload malicious files by sending ZIP archives through the ft2.php endpoint. Attackers can upload ZIP files containing PHP shells, use the unzip functionality to extract them into accessible directories, and execute arbitrary commands through the extracted PHP files.

leefish

Craft is a content management system (CMS). The ElementSearchController::actionSearch() endpoint is missing the unset() protection that was added to ElementIndexesController in CVE-2026-25495. The exact same SQL injection vulnerability (including criteria[orderBy], the original advisory vector) works on this controller because the fix was never applied to it. Any authenticated control panel user (no admin required) can inject arbitrary SQL via criteria[where], criteria[orderBy], or other query properties, and extract the full database contents via boolean-based blind injection. Users should update to the patched 5.9.9 release to mitigate the issue.

craftcms

In Splunk Enterprise versions below 10.2.1, 10.0.4, 9.4.9, and 9.3.10, and Splunk Cloud Platform versions below 10.2.2510.7, 10.1.2507.17, 10.0.2503.12, and 9.3.2411.124, a low-privileged user that does not hold the "admin" or "power" Splunk roles could retrieve sensitive information by inspecting the job's search log due to improper access control in the MongoClient logging channel.

splunk

In Splunk Enterprise versions below 10.2.0, 10.0.4, 9.4.9, and 9.3.10, and Splunk Cloud Platform versions below 10.2.2510.5, 10.0.2503.12, 10.1.2507.16, and 9.3.2411.124, a user who holds a role that contains the high-privilege capability `edit_cmd` could execute arbitrary shell commands using the `unarchive_cmd` parameter for the `/splunkd/__upload/indexing/preview` REST endpoint.

splunk

OpenProject is an open-source, web-based project management software. Prior to 17.2.0, an authenticated project member with BCF import permissions can upload a crafted .bcf archive where the <Snapshot> value in markup.bcf is manipulated to contain an absolute or traversal local path (for example: /etc/passwd or ../../../../etc/passwd). During import, this untrusted <Snapshot> value is used as file.path during attachment processing. As a result, local filesystem content can be read outside the intended ZIP scope. This results in an Arbitrary File Read (AFR) within the read permissions of the OpenProject application user. This vulnerability is fixed in 17.2.0.

openproject

GitLab has remediated an issue in GitLab CE/EE affecting all versions from 10.0 before 18.7.6, 18.8 before 18.8.6, and 18.9 before 18.9.2 that could have allowed an unauthenticated user to cause a denial of service by issuing specially crafted requests to repository archive endpoints under certain conditions.

gitlab

The Gravity Forms plugin for WordPress is vulnerable to Stored Cross-Site Scripting in all versions up to, and including, 2.9.28.1. This is due to a compound failure involving missing authorization on the `create_from_template` AJAX endpoint (allowing any authenticated user to create forms), insufficient input sanitization (`sanitize_text_field()` preserves single quotes), and missing output escaping when the form title is rendered in the Form Switcher dropdown (`title` attribute constructed without `esc_attr()`, and JavaScript `saferHtml` utility only escapes `&`, `<`, `>` but not quotes). This makes it possible for authenticated attackers, with Subscriber-level access and above, to inject arbitrary JavaScript that executes when an Administrator searches in the Form Switcher dropdown in the Form Editor.

Illustrator versions 29.8.4, 30.1 and earlier are affected by an Untrusted Search Path vulnerability that might allow attackers to execute arbitrary code in the context of the current user. Exploitation of this issue requires user interaction in that a victim must open a malicious file.

adobe · microsoft

Parse Server is an open source backend that can be deployed to any infrastructure that can run Node.js. Prior to 9.5.2-alpha.13 and 8.6.26, the LDAP authentication adapter is vulnerable to LDAP injection. User-supplied input (authData.id) is interpolated directly into LDAP Distinguished Names (DN) and group search filters without escaping special characters. This allows an attacker with valid LDAP credentials to manipulate the bind DN structure and to bypass group membership checks. This enables privilege escalation from any authenticated LDAP user to a member of any restricted group. The vulnerability affects Parse Server deployments that use the LDAP authentication adapter with group-based access control. This vulnerability is fixed in 9.5.2-alpha.13 and 8.6.26.

parseplatform

LinkAce is a self-hosted archive to collect website links. In 2.1.0 and earlier, the processTaxonomy() method in LinkRepository.php allows authenticated users to attach other users' private tags and lists to their own links by passing integer IDs.

linkace

LinkAce is a self-hosted archive to collect website links. When a user creates a link via POST /links, the server fetches HTML metadata from the provided URL (LinkRepository::create() calls HtmlMeta::getFromUrl()). The LinkStoreRequest validation rules do not include NoPrivateIpRule, allowing server-side requests to internal network addresses, Docker service hostnames, and cloud metadata endpoints. The project already has a NoPrivateIpRule class (app/Rules/NoPrivateIpRule.php) but it is only applied in FetchController.php (line 99), not in the primary link creation path.

linkace