CVE Vulnerability Catalog

Tandoor Recipes is an application for managing recipes, planning meals, and building shopping lists. Prior to 2.6.5, a critical Denial of Service (DoS) vulnerability existed in the recipe import functionality, allowing an authenticated user to crash the server or significantly degrade its performance by uploading a large ZIP file (ZIP Bomb); this is fixed in 2.6.5.

Chamilo LMS prior to 2.0.0-RC.3 contains an Insecure Direct Object Reference (IDOR) vulnerability in the REST API stats endpoint. Any authenticated user (including low-privilege students with ROLE_USER) can read any other user's learning progress, certificates, and gradebook scores for any course, without enrollment or supervisory relationship.

Chamilo LMS prior to 1.11.38 and 2.0.0-RC.3 contains an Open Redirect vulnerability in the session course edit page. An attacker can redirect an authenticated administrator to an arbitrary external URL after saving coach assignment changes, and the redirect leaks the id_session parameter to the attacker's server.

Chamilo LMS prior to 2.0.0-RC.3 contains a Reflected Cross-Site Scripting (XSS) vulnerability in the exercise question list admin panel. The pagination code merges all $_GET parameters via array_merge() and outputs the result via http_build_query() directly into HTML href attributes without htmlspecialchars() encoding, allowing execution of arbitrary JavaScript in an authenticated teacher's browser.

A vulnerability in the HTTP client allows CR/LF bytes to be injected into proxy tunnel headers or host, potentially leading to HTTP response splitting or cache poisoning attacks.

PraisonAIAgents is a multi-agent teams system. Prior to 1.5.128, web_crawl's httpx fallback path passes user-supplied URLs directly to httpx.AsyncClient.get() with follow_redirects=True and no host validation. An LLM agent tricked into crawling an internal URL can reach cloud metadata endpoints (169.254.169.254), internal services, and localhost, and the response content is returned to the agent and may appear in output visible to the attacker.

PraisonAI prior to 4.5.128 allows spawning background servers via stdio using user-supplied command strings in its MCP integration. By default, the entire parent process environment is forwarded to the spawned subprocess, exposing sensitive data such as API keys, authentication tokens, and database credentials.

In Vikunja prior to 2.3.0, scoped API token enforcement for custom project background routes is method-confused. A token with only projects.background can delete a project background, while a token with only projects.background_delete is rejected, constituting a scoped-token authorization bypass.

FastGPT prior to 4.14.10.3 has an endpoint /api/core/app/mcpTools/runTool that accepts arbitrary URLs without authentication. The internal IP check in isInternalAddress() only blocks private IPs when CHECK_INTERNAL_IP=true, which is not the default. This allows unauthenticated attackers to perform SSRF against internal network resources.

In rembg before 2.0.75, a path traversal vulnerability in the HTTP server allows unauthenticated remote attackers to read arbitrary files from the filesystem. By sending a crafted model_path parameter, an attacker can force the server to attempt loading any file as an ONNX model, revealing file existence, permissions, and potentially contents.

In Vikunja prior to 2.3.0, the file import endpoint uses the attacker-controlled Size field from the JSON metadata inside the import zip instead of the actual decompressed file content length for the file size enforcement check. Setting Size to 0 while including large compressed file entries in the zip bypasses the configured maximum file size limit.

Vikunja prior to 2.3.0 builds iCalendar VTODO entries via raw string concatenation without RFC 5545 TEXT escaping. User-controlled task titles containing CRLF can break property boundaries and inject arbitrary iCalendar properties such as ATTACH, VALARM, or ORGANIZER.

In Vikunja prior to 2.3.0, task titles are embedded directly into Markdown link syntax in overdue email notifications without escaping Markdown special characters. When rendered by goldmark and sanitized by bluemonday (which allows <a> and <img> tags), injected Markdown constructs produce phishing links and tracking pixels in legitimate notification emails. This vulnerability is fixed in 2.3.0.

In Vikunja prior to 2.3.0, the addRepeatIntervalToTime function uses an O(n) loop that advances a date by the task's RepeatAfter duration until it exceeds the current time. Creating a repeating task with a 1-second interval and a due date far in the past triggers billions of loop iterations, consuming CPU and holding a database connection for minutes per request.

In Vikunja prior to 2.3.0, the CalDAV GetResource and GetResourcesByList methods fetch tasks by UID from the database without verifying that the authenticated user has access to the task's project. Any authenticated CalDAV user who knows a task UID can read full task data from any project.

Vikunja is an open-source self-hosted task management platform. Prior to 2.3.0, the TOTP failed-attempt lockout mechanism was non-functional due to a database transaction handling bug: after a failed TOTP validation, the login handler calls HandleFailedTOTPAuth and then unconditionally rolls back, undoing the StatusAccountLocked write and allowing unlimited brute-force attempts against TOTP codes; this is fixed in 2.3.0.

Vikunja prior to 2.3.0 contains a SQL operator precedence bug in the hasAccessToLabel function that allows any authenticated user to read any label with at least one task association, regardless of project access. Label titles, descriptions, colors, and creator information are exposed. The vulnerability is fixed in 2.3.0.

Apache Log4cxx before 1.7.0 (XMLLayout) fails to sanitize characters forbidden by the XML 1.0 specification in log messages and NDC/MDC property keys and values, producing invalid XML. Conforming XML parsers reject such documents with a fatal error, which may cause downstream log processing systems to drop or fail to index affected records. An attacker who can influence logged data can suppress individual log records, impairing audit trails.

Apache Log4net before 3.3.0, in XmlLayout and XmlLayoutSchemaLog4J, fails to sanitize characters forbidden by the XML 1.0 specification in MDC property keys and values, as well as the identity field that may carry attacker-influenced data. This causes an exception during serialization and silent loss of the affected log event.

Vikunja is an open-source self-hosted task management platform. Prior to 2.3.0, link share authentication (GetLinkShareFromClaims in pkg/models/link_sharing.go) constructs authorization objects entirely from JWT claims without server-side database validation. When a project owner deletes a link share or downgrades its permissions, previously issued JWTs continue to grant the original permission level for up to 72 hours.