Firecrawl turns entire websites into LLM-ready markdown or structured data. Prior to version 2.0.1, a server-side request forgery (SSRF) vulnerability was discovered in Firecrawl's webhook functionality. Authenticated users could configure a webhook to an internal URL and send POST requests with arbitrary headers, which may have allowed access to internal systems. This has been fixed in version 2.0.1. If upgrading is not possible, it is recommend to isolate Firecrawl from any sensitive internal systems.

ImageMagick is free and open-source software used for editing and manipulating digital images. Prior to versions 6.9.13-28 and 7.1.2-2 for ImageMagick's 32-bit build, a 32-bit integer overflow in the BMP encoder’s scanline-stride computation collapses bytes_per_line (stride) to a tiny value while the per-row writer still emits 3 × width bytes for 24-bpp images. The row base pointer advances using the (overflowed) stride, so the first row immediately writes past its slot and into adjacent heap memory with attacker-controlled bytes. This is a classic, powerful primitive for heap corruption in common auto-convert pipelines. This issue has been patched in versions 6.9.13-28 and 7.1.2-2.

ImageMagick is free and open-source software used for editing and manipulating digital images. Prior to ImageMagick versions 6.9.13-28 and 7.1.2-2, a format string bug vulnerability exists in InterpretImageFilename function where user input is directly passed to FormatLocaleString without proper sanitization. An attacker can overwrite arbitrary memory regions, enabling a wide range of attacks from heap overflow to remote code execution. This issue has been patched in versions 6.9.13-28 and 7.1.2-2.

ImageMagick is free and open-source software used for editing and manipulating digital images. Prior to versions 6.9.13-28 and 7.1.2-2, passing a geometry string containing only a colon (":") to montage -geometry leads GetGeometry() to set width/height to 0. Later, ThumbnailImage() divides by these zero dimensions, triggering a crash (SIGFPE/abort), resulting in a denial of service. This issue has been patched in versions 6.9.13-28 and 7.1.2-2.

traQ is a messenger application built for Digital Creators Club traP. Prior to version 3.25.0, a vulnerability exists where sensitive information, such as OAuth tokens, are recorded in log files when an error occurs during the execution of an SQL query. An attacker could intentionally trigger an SQL error by methods such as placing a high load on the database. This could allow an attacker who has the authority to view the log files to illicitly acquire the recorded sensitive information. This vulnerability has been patched in version 3.25.0. If upgrading is not possible, a temporary workaround involves reviewing access permissions for SQL error logs and strictly limiting access to prevent unauthorized users from viewing them.

jsPDF is a library to generate PDFs in JavaScript. Prior to 3.0.2, user control of the first argument of the addImage method results in CPU utilization and denial of service. If given the possibility to pass unsanitized image data or URLs to the addImage method, a user can provide a harmful PNG file that results in high CPU utilization and denial of service. The vulnerability was fixed in jsPDF 3.0.2.

request-filtering-agent is an http(s).Agent implementation that blocks requests to Private/Reserved IP addresses. Versions 1.x.x and earlier contain a vulnerability where HTTPS requests to 127.0.0.1 bypass IP address filtering, while HTTP requests are correctly blocked. This allows attackers to potentially access internal HTTPS services running on localhost, bypassing the library's SSRF protection. The vulnerability is particularly dangerous when the application accepts user-controlled URLs and internal services are only protected by network-level restrictions. This vulnerability has been fixed in request-filtering-agent version 2.0.0. Users should upgrade to version 2.0.0 or later.

XGrammar is an open-source library for efficient, flexible, and portable structured generation. Prior to version 0.1.21, XGrammar has an infinite recursion issue in the grammar. This issue has been resolved in version 0.1.21.

The Scratch Channel is a news website. In versions 1 and 1.1, a POST request to the endpoint used to publish articles, can be used to post an article in any category with any date, regardless of who's logged in. This issue has been patched in version 1.2.

h2 is a pure-Python implementation of a HTTP/2 protocol stack. Prior to version 4.3.0, an HTTP/2 request splitting vulnerability allows attackers to perform request smuggling attacks by injecting CRLF characters into headers. This occurs when servers downgrade HTTP/2 requests to HTTP/1.1 without properly validating header names/values, enabling attackers to manipulate request boundaries and bypass security controls. This issue has been patched in version 4.3.0.

Craft is a platform for creating digital experiences. From versions 4.0.0-RC1 to 4.16.5 and 5.0.0-RC1 to 5.8.6, there is a potential remote code execution vulnerability via Twig SSTI (Server-Side Template Injection). This is a follow-up to CVE-2024-52293. This vulnerability has been patched in versions 4.16.6 and 5.8.7.

Airlink's Daemon interfaces with Docker and the Panel to provide secure access for controlling instances via the Panel. In version 1.0.0, an attacker with access to the affected container can create symbolic links inside the mounted directory (/app/data). Because the container bind-mounts an arbitrary host path, these symlinks can point to sensitive locations on the host filesystem. When the application or other processes follow these symlinks, the attacker can gain unauthorized read access to host files outside the container. This issue has been patched in version 1.0.1.

DataEase is an open source business intelligence and data visualization tool. Prior to version 2.10.12, there is a H2 JDBC RCE bypass in DataEase. If the JDBC URL meets criteria, the getJdbcUrl method is returned, which acts as the getter for the JdbcUrl parameter provided. This bypasses H2's filtering logic and returns the H2 JDBC URL, allowing the "driver":"org.h2.Driver" to specify the H2 driver for the JDBC connection. The vulnerability has been fixed in version 2.10.12.

DataEase is an open source business intelligence and data visualization tool. Prior to version 2.10.12, because DB2 parameters are not filtered, a JNDI injection attack can be directly launched. JNDI triggers an AspectJWeaver deserialization attack, writing to various files. This vulnerability requires commons-collections 4.x and aspectjweaver-1.9.22.jar. The vulnerability has been fixed in version 2.10.12.

Langflow is a tool for building and deploying AI-powered agents and workflows. A privilege escalation vulnerability exists in Langflow containers where an authenticated user with RCE access can invoke the internal CLI command langflow superuser to create a new administrative user. This results in full superuser access, even if the user initially registered through the UI as a regular (non-admin) account. A patched version has not been made public at this time.

The Scratch Channel is a news website. In version 1, it is possible to go to application in devtools and click local storage to edit the account's username locally. This issue has been patched in version 1.1.

PhpOffice/PhpSpreadsheet is a pure PHP library for reading and writing spreadsheet files. Prior to versions 1.30.0, 2.1.12, 2.4.0, 3.10.0, and 5.0.0, SSRF can occur when a processed HTML document is read and displayed in the browser. The vulnerability lies in the setPath method of the PhpOffice\PhpSpreadsheet\Worksheet\Drawing class, where a crafted string from the user is passed to the HTML reader. This issue has been patched in versions 1.30.0, 2.1.12, 2.4.0, 3.10.0, and 5.0.0.

gnark is a zero-knowledge proof system framework. In versions prior to 0.14.0, the Verify function in eddsa.go and ecdsa.go used the S value from a signature without asserting that 0 ≤ S < order, leading to a signature malleability vulnerability. Because gnark’s native EdDSA and ECDSA circuits lack essential constraints, multiple distinct witnesses can satisfy the same public inputs. In protocols where nullifiers or anti-replay checks are derived from R and S, this enables signature malleability and may allow double spending. This issue has been addressed in version 0.14.0.

Audiobookshelf is an open-source self-hosted audiobook server. In versions 2.6.0 through 2.26.3, the application does not properly restrict redirect callback URLs during OIDC authentication. An attacker can craft a login link that causes Audiobookshelf to store an arbitrary callback in a cookie, which is later used to redirect the user after authentication. The server then issues a 302 redirect to the attacker-controlled URL, appending sensitive OIDC tokens as query parameters. This allows an attacker to obtain the victim's tokens and perform full account takeover, including creating persistent admin users if the victim is an administrator. Tokens are further leaked via browser history, Referer headers, and server logs. This vulnerability impacts all Audiobookshelf deployments using OIDC; no IdP misconfiguration is required. The issue is fixed in version 2.28.0. No known workarounds exist.

The open-source identity infrastructure software Zitadel allows administrators to disable the user self-registration. Versions 4.0.0 to 4.0.2, 3.0.0 to 3.3.6, and all versions prior to 2.71.15 are vulnerable to a username enumeration issue in the login interface. The login UI includes a security feature, Ignoring unknown usernames, that is intended to prevent username enumeration by returning a generic response for both valid and invalid usernames. This vulnerability allows an unauthenticated attacker to bypass this protection by submitting arbitrary userIDs to the select account page and distinguishing between valid and invalid accounts based on the system's response. For effective exploitation, an attacker needs to iterate through possible userIDs, but the impact can be limited by implementing rate limiting or similar measures. The issue has been patched in versions 4.0.3, 3.4.0, and 2.71.15.

Roo Code is an AI-powered autonomous coding agent that lives in users' editors. In versions prior to 3.25.5, Roo-Code fails to properly handle process substitution and single ampersand characters in the command parsing logic for auto-execute commands. If a user has enabled auto-approved execution for a command such as ls, an attacker who can submit crafted prompts to the agent may inject arbitrary commands to be executed alongside the intended command. Exploitation requires attacker access to submit prompts and for the user to have enabled auto-approved command execution, which is disabled by default. This vulnerability could allow an attacker to execute arbitrary code. The issue is fixed in version 3.25.5.

UnoPim is an open-source Product Information Management (PIM) system built on the Laravel framework. Versions 0.3.0 and prior are vulnerable to CSV injection, also known as formula injection, in the Quick Export feature. This vulnerability allows attackers to inject malicious content into exported CSV files. When the CSV file is opened in spreadsheet applications such as Microsoft Excel, the malicious input may be interpreted as a formula or command, potentially resulting in the execution of arbitrary code on the victim's device. Successful exploitation can lead to remote code execution, including the establishment of a reverse shell. Users are advised to upgrade to version 0.3.1 or later.

dpanel is an open source server management panel written in Go. In versions 1.2.0 through 1.7.2, dpanel allows authenticated users to read arbitrary files from the server via the /api/app/compose/get-from-uri API endpoint. The vulnerability exists in the GetFromUri function in app/application/http/controller/compose.go, where the uri parameter is passed directly to os.ReadFile without proper validation or access control. A logged-in attacker can exploit this flaw to read sensitive files from the host system, leading to information disclosure. No patched version is available as of this writing.

pyLoad is the free and open-source Download Manager written in pure Python. The jk parameter is received in pyLoad CNL Blueprint. Due to the lack of jk parameter verification, the jk parameter input by the user is directly determined as dykpy.evaljs(), resulting in the server CPU being fully occupied and the web-ui becoming unresponsive. This vulnerability is fixed in 0.5.0b3.dev92.

Phproject is a high performance full-featured project management system. From 1.8.0 to before 1.8.3, a Stored Cross-Site Scripting (XSS) vulnerability exists in the Planned Hours field when creating a new project. When sending a POST request to /issues/new/, the value provided in the Planned Hours field is included in the server response without any HTML encoding or sanitization. Because of this, an attacker can craft a malicious payload such as <script>alert(1)</script> and include it in the planned_hours parameter. The server reflects the input directly in the HTML of the project creation page, causing the browser to interpret and execute it. This vulnerability is fixed in 1.8.3.

WeGIA is a Web manager for charitable institutions. Prior to 3.4.7, a Reflected Cross-Site Scripting (XSS) vulnerability was identified in the pre_cadastro_adotante.php endpoint of the WeGIA application. This vulnerability allows attackers to inject malicious scripts in the msg_e parameter. This vulnerability is fixed in 3.4.7.

WeGIA is a Web manager for charitable institutions. Prior to 3.4.7, a Reflected Cross-Site Scripting (XSS) vulnerability was identified in the cargos.php endpoint of the WeGIA application. This vulnerability allows attackers to inject malicious scripts in the msg_e parameter. This vulnerability is fixed in 3.4.7.

WeGIA is a Web manager for charitable institutions. Prior to 3.4.7, there is a Reflected Cross-Site Scripting (XSS) vulnerability in the insere_despacho.php endpoint of the WeGIA application. This vulnerability allows attackers to inject malicious scripts in the cpf sccs. This vulnerability is fixed in 3.4.7.

WeGIA is a Web manager for charitable institutions. Prior to 3.4.7, there is a Stored Cross-Site Scripting (XSS) vulnerability in the dependente_docdependente.php endpoint of the WeGIA application. This vulnerability allows attackers to inject malicious scripts into the nome parameter. The injected scripts are stored on the server and executed automatically whenever the affected page is accessed by users, posing a significant security risk. This vulnerability is fixed in 3.4.7.

WeGIA is a Web manager for charitable institutions. Prior to 3.4.10, there is a SQL Injection vulnerability in the /html/funcionario/dependente_remover.php endpoint, specifically in the id_funcionario parameter. This vulnerability allows attackers to execute arbitrary SQL commands, compromising the confidentiality, integrity, and availability of the database. This vulnerability is fixed in 3.4.10.

claude-code-router is a powerful tool to route Claude Code requests to different models and customize any request. Due to improper Cross-Origin Resource Sharing (CORS) configuration, there is a risk that user API Keys or equivalent credentials may be exposed to untrusted domains. Attackers could exploit this misconfiguration to steal credentials, abuse accounts, exhaust quotas, or access sensitive data. The issue has been patched in v1.0.34.

eslint-ban-moment is an Eslint plugin for final assignment in VIHU. In 3.0.0 and earlier, a sensitive Supabase URI is exposed in .env. A valid Supabase URI with embedded username and password will allow an attacker complete unauthorized access and control over database and user data. This could lead to data exfiltration, modification or deletion.

vite-plugin-static-copy is rollup-plugin-copy for Vite with dev server support. Files not included in src are accessible with a crafted request. The vulnerability is fixed in 2.3.2 and 3.1.2.

UnoPim is an open-source Product Information Management (PIM) system built on the Laravel framework. Before 0.2.1, some of the endpoints of the application is vulnerable to Cross site Request forgery (CSRF). This vulnerability is fixed in 0.2.1.

UnoPim is an open-source Product Information Management (PIM) system built on the Laravel framework. Before 0.2.1, the image upload at the user creation feature performs only client side file type validation. A user can capture the request by uploading an image, capture the request through a Proxy like Burp suite. Make changes to the file extension and content. The vulnerability is fixed in 0.2.1.

UnoPim is an open-source Product Information Management (PIM) system built on the Laravel framework. Before 0.2.1, UnoPim contains a stored cross-site scripting vulnerability via SVG MIME/sanitizer bypass in the /admin/settings/users/create endpoint. This vulnerability is fixed in 0.2.1.

ESF-IDF is the Espressif Internet of Things (IOT) Development Framework. The BluFi example bundled in ESP-IDF was vulnerable to memory overflows in two areas: Wi-Fi credential handling and Diffie–Hellman key exchange. This vulnerability is fixed in 5.4.1, 5.3.3, 5.1.6, and 5.0.9.

vLLM is an inference and serving engine for large language models (LLMs). From 0.1.0 to before 0.10.1.1, a Denial of Service (DoS) vulnerability can be triggered by sending a single HTTP GET request with an extremely large header to an HTTP endpoint. This results in server memory exhaustion, potentially leading to a crash or unresponsiveness. The attack does not require authentication, making it exploitable by any remote user. This vulnerability is fixed in 0.10.1.1.

n8n is a workflow automation platform. Before 1.106.0, a symlink traversal vulnerability was discovered in the Read/Write File node in n8n. While the node attempts to restrict access to sensitive directories and files, it does not properly account for symbolic links (symlinks). An attacker with the ability to create symlinks—such as by using the Execute Command node—could exploit this to bypass the intended directory restrictions and read from or write to otherwise inaccessible paths. Users of n8n.cloud are not impacted. Affected users should update to version 1.106.0 or later.

Directus is a real-time API and App dashboard for managing SQL database content. From 10.8.0 to before 11.9.3, a vulnerability exists in the file update mechanism which allows an unauthenticated actor to modify existing files with arbitrary contents (without changes being applied to the files' database-resident metadata) and / or upload new files, with arbitrary content and extensions, which won't show up in the Directus UI. This vulnerability is fixed in 11.9.3.

OnboardLite is the result of the Influx Initiative, our vision for an improved student organization lifecycle at the University of Central Florida. An attacker can craft a link to the trusted application that, when visited, redirects the user to a malicious external site. This enables phishing, credential theft, malware delivery, and trust abuse. Any version with commit hash 6cca19e or later implements jwt signing for the redirect url parameter.

Frappe is a full-stack web application framework. Prior to 15.74.2 and 14.96.15, an attacker could implement SQL injection through specially crafted requests, allowing malicious people to access sensitive information. This vulnerability is a bypass of the official patch released for CVE-2025-52895. This vulnerability is fixed in 15.74.2 and 14.96.15.

Frappe is a full-stack web application framework. A carefully crafted request could extract data that the user would normally not have access to, via SQL injection. This vulnerability is fixed in 15.74.2 and 14.96.15.

nginx-defender is a high-performance, enterprise-grade Web Application Firewall (WAF) and threat detection system engineered for modern web infrastructure. This is a configuration vulnerability affecting nginx-defender deployments. Example configuration files config.yaml and docker-compose.yml contain default credentials (default_password: "change_me_please", GF_SECURITY_ADMIN_PASSWORD=admin123). If users deploy nginx-defender without changing these defaults, attackers with network access could gain administrative control, bypassing security protections. The issue is addressed in v1.5.0 and later.

flaskBlog is a blog app built with Flask. In 2.8.0 and earlier, when deleting a comment, there's no validation of the ownership of the comment. Every user can delete an arbitrary comment of another user on every post, by simply intercepting the delete request and changing the commentID. The code that causes the problem is in routes/post.py.

flaskBlog is a blog app built with Flask. In 2.8.0 and earlier, an arbitrary user can change his role to "admin", giving its relative privileges (e.g. delete users, posts, comments etc.). The problem is in the routes/adminPanelUsers file.

flaskBlog is a blog app built with Flask. In 2.8.0 and earlier, when creating a post, there's no validation of the content of the post stored in the variable "postContent". The vulnerability arises when displaying the content of the post using the | safe filter, that tells the engine to not escape the rendered content. This can lead to a stored XSS inside the content of the post. The code that causes the problem is in template/routes.html.

flaskBlog is a blog app built with Flask. In 2.8.0 and earlier, the code checks if the userRole is "admin" only when visiting the /admin page, but not when visiting its subroutes. Specifically, only the file routes/adminPanel.py checks the user role when a user is trying to access the admin page, but that control is not done for the pages routes/adminPanelComments.py and routes/adminPanelPosts.py. Thus, an unauthorized user can bypass the intended restrictions, leaking sensitive data and accessing the following pages: /admin/posts, /adminpanel/posts, /admin/comments, and /adminpanel/comments.

DeepChat is a smart assistant that connects powerful AI to your personal world. DeepChat before 0.3.1 has a one-click remote code execution vulnerability. An attacker can exploit this vulnerability by embedding a specially crafted deepchat: URL on any website, including a malicious one they control. When a victim visits such a site or clicks on the link, the browser triggers the app’s custom URL handler (deepchat:), causing the DeepChat application to launch and process the URL, leading to remote code execution on the victim’s machine. This vulnerability is fixed in 0.3.1.