CVE Database

PraisonAI is a multi-agent teams system. Prior to 4.5.128, cmd_unpack in the recipe CLI extracts .praison tar archives using raw tar.extract() without validating archive member paths. A .praison bundle containing ../../ entries will write files outside the intended output directory. An attacker who distributes a malicious bundle can overwrite arbitrary files on the victim's filesystem when they run praisonai recipe unpack. This vulnerability is fixed in 4.5.128.

PraisonAI is a multi-agent teams system. Prior to 4.5.128, PraisonAI automatically loads a file named tools.py from the current working directory to discover and register custom agent tools. This loading process uses importlib.util.spec_from_file_location and immediately executes module-level code via spec.loader.exec_module() without explicit user consent, validation, or sandboxing. The tools.py file is loaded implicitly, even when it is not referenced in configuration files or explicitly requested by the user. As a result, merely placing a file named tools.py in the working directory is sufficient to trigger code execution. This behavior violates the expected security boundary between user-controlled project files (e.g., YAML configurations) and executable code, as untrusted content in the working directory is treated as trusted and executed automatically. If an attacker can place a malicious tools.py file into a directory where a user or automated system (e.g., CI/CD pipeline) runs praisonai, arbitrary code execution occurs immediately upon startup, before any agent logic begins. This vulnerability is fixed in 4.5.128.

SvelteKit is a framework for rapidly developing robust, performant web applications using Svelte. Prior to 2.57.1, redirect, when called from inside the handle server hook with a location parameter containing characters that are invalid in a HTTP header, will cause an unhandled TypeError. This could result in DoS on some platforms, especially if the location passed to redirect contains unsanitized user input. This vulnerability is fixed in 2.57.1.

SvelteKit is a framework for rapidly developing robust, performant web applications using Svelte. Prior to 2.57.1, under certain circumstances, requests could bypass the BODY_SIZE_LIMIT on SvelteKit applications running with adapter-node. This bypass does not affect body size limits at other layers of the application stack, so limits enforced in the WAF, gateway, or at the platform level are unaffected. This vulnerability is fixed in 2.57.1.

OpenClaw before 2026.3.25 contains a privilege escalation vulnerability in gateway-authenticated plugin HTTP routes that incorrectly mint operator.admin runtime scope regardless of caller-granted scopes. Attackers can exploit this scope boundary bypass to gain elevated privileges and perform unauthorized administrative actions.

OpenClaw before 2026.3.24 contains a path traversal vulnerability in sandbox enforcement allowing sandboxed agents to read arbitrary files from other agents' workspaces via unnormalized mediaUrl or fileUrl parameter keys. Attackers can exploit incomplete parameter validation in normalizeSandboxMediaParams and missing mediaLocalRoots context to access sensitive files including API keys and configuration data outside designated sandbox roots.

OpenClaw before 2026.3.22 contains an allowlist bypass vulnerability in system.run approvals that fails to unwrap /usr/bin/time wrappers. Attackers can bypass executable binding restrictions by using an unregistered time wrapper to reuse approval state for inner commands.

OpenClaw before 2026.3.25 contains a privilege escalation vulnerability allowing non-admin operators to self-request broader scopes during backend reconnect. Attackers can bypass pairing requirements to reconnect as operator.admin, gaining unauthorized administrative privileges.

OpenClaw before 2026.3.23 contains an insufficient access control vulnerability in the Gateway agent /reset endpoint that allows callers with operator.write permission to reset admin sessions. Attackers with operator.write privileges can invoke /reset or /new messages with an explicit sessionKey to bypass operator.admin requirements and reset arbitrary sessions.

OpenClaw before 2026.3.24 contains an incorrect authorization vulnerability in the POST /reset-profile endpoint that allows authenticated callers with operator.write access to browser.request to bypass profile mutation restrictions. Attackers can invoke POST /reset-profile through the browser.request surface to stop the running browser, close Playwright connections, and move profile directories to Trash, crossing intended privilege boundaries.

OpenClaw before 2026.3.22 contains an environment variable override handling vulnerability that allows attackers to bypass the shared host environment policy through inconsistent sanitization paths. Attackers can supply blocked or malformed override keys that slip through inconsistent validation to execute arbitrary code with unintended environment variables.

OpenClaw before 2026.3.22 contains an unvalidated WebView JavascriptInterface vulnerability allowing attackers to inject arbitrary instructions. Untrusted pages can invoke the canvas bridge to execute malicious code within the Android application context.

OpenClaw before 2026.3.24 contains an arbitrary code execution vulnerability in local plugin and hook installation that allows attackers to execute malicious code by crafting a .npmrc file with a git executable override. During npm install execution in the staged package directory, attackers can leverage git dependencies to trigger execution of arbitrary programs specified in the attacker-controlled .npmrc configuration file.

Vikunja is an open-source self-hosted task management platform. Prior to 2.3.0, the CanUpdate check at pkg/models/project_permissions.go:139-148 only requires CanWrite on the new parent project when changing parent_project_id. However, Vikunja's permission model uses a recursive CTE that walks up the project hierarchy to compute permissions. Moving a project under a different parent changes the permission inheritance chain. When a user has inherited Write access (from a parent project share) and reparents the child project under their own project tree, the CTE resolves their ownership of the new parent as Admin (permission level 2) on the moved project. This vulnerability is fixed in 2.3.0.

Vikunja is an open-source self-hosted task management platform. Prior to 2.3.0, the OIDC callback handler issues a full JWT token without checking whether the matched user has TOTP two-factor authentication enabled. When a local user with TOTP enrolled is matched via the OIDC email fallback mechanism, the second factor is completely skipped. This vulnerability is fixed in 2.3.0.

Apache Log4j's JsonTemplateLayout https://logging.apache.org/log4j/2.x/manual/json-template-layout.html , in versions up to and including 2.25.3, produces invalid JSON output when log events contain non-finite floating-point values (NaN, Infinity, or -Infinity), which are prohibited by RFC 8259. This may cause downstream log processing systems to reject or fail to index affected records. An attacker can exploit this issue only if both of the following conditions are met: * The application uses JsonTemplateLayout. * The application logs a MapMessage, or logs an object directly (e.g., via Logger.info(Object), which wraps it in an ObjectMessage), where the message contains an attacker-controlled floating-point value. Users are advised to upgrade to Apache Log4j JSON Template Layout 2.25.4, which corrects this issue. Note: The fix released in version 2.25.4 did not cover all affected code paths. CVE-2026-49844 was assigned to the remaining issue, which concerns the MapMessage.asJson() serialization in Apache Log4j API and is fixed in versions 2.25.5 and 2.26.1.

Apache Log4j Core's XmlLayout https://logging.apache.org/log4j/2.x/manual/layouts.html#XmlLayout , in versions up to and including 2.25.3, fails to sanitize characters forbidden by the XML 1.0 specification https://www.w3.org/TR/xml/#charsets producing invalid XML output whenever a log message or MDC value contains such characters. The impact depends on the StAX implementation in use: * JRE built-in StAX: Forbidden characters are silently written to the output, producing malformed XML. Conforming parsers must reject such documents with a fatal error, which may cause downstream log-processing systems to drop the affected records. * Alternative StAX implementations (e.g., Woodstox https://github.com/FasterXML/woodstox , a transitive dependency of the Jackson XML Dataformat module): An exception is thrown during the logging call, and the log event is never delivered to its intended appender, only to Log4j's internal status logger. Users are advised to upgrade to Apache Log4j Core 2.25.4, which corrects this issue by sanitizing forbidden characters before XML output.

The Log4j1XmlLayout from the Apache Log4j 1-to-Log4j 2 bridge fails to escape characters forbidden by the XML 1.0 standard, producing malformed XML output. Conforming XML parsers are required to reject documents containing such characters with a fatal error, which may cause downstream log processing systems to drop or fail to index affected records. Two groups of users are affected: * Those using Log4j1XmlLayout directly in a Log4j Core 2 configuration file. * Those using the Log4j 1 configuration compatibility layer with org.apache.log4j.xml.XMLLayout specified as the layout class. Users are advised to upgrade to Apache Log4j 1-to-Log4j 2 bridge version 2.25.4, which corrects this issue. Note: The Apache Log4j 1-to-Log4j 2 bridge is deprecated and will not be present in Log4j 3. Users are encouraged to consult the Log4j 1 to Log4j 2 migration guide https://logging.apache.org/log4j/2.x/migrate-from-log4j1.html , and specifically the section on eliminating reliance on the bridge.

Apache Log4j Core's Rfc5424Layout https://logging.apache.org/log4j/2.x/manual/layouts.html#RFC5424Layout , in versions 2.21.0 through 2.25.3, is vulnerable to log injection via CRLF sequences due to undocumented renames of security-relevant configuration attributes. Two distinct issues affect users of stream-based syslog services who configure Rfc5424Layout directly: * The newLineEscape attribute was silently renamed, causing newline escaping to stop working for users of TCP framing (RFC 6587), exposing them to CRLF injection in log output. * The useTlsMessageFormat attribute was silently renamed, causing users of TLS framing (RFC 5425) to be silently downgraded to unframed TCP (RFC 6587), without newline escaping. Users of the SyslogAppender are not affected, as its configuration attributes were not modified. Users are advised to upgrade to Apache Log4j Core 2.25.4, which corrects this issue.

CouchCMS contains a privilege escalation vulnerability that allows authenticated Admin-level users to create SuperAdmin accounts by tampering with the f_k_levels_list parameter in user creation requests. Attackers can modify the parameter value from 4 to 10 in the HTTP request body to bypass authorization validation and gain full application control, circumventing restrictions on SuperAdmin account creation and privilege assignment.

An issue was discovered in BMC Control-M/MFT 9.0.20 through 9.0.22. An API management endpoint allows unauthenticated users to obtain both an API identifier and its corresponding secret value. With these exposed secrets, an attacker could invoke privileged API operations, potentially leading to unauthorized access.

An issue was discovered in BMC Control-M/MFT 9.0.20 through 9.0.22. A SQL injection vulnerability in the MFT API's debug interface allows an authenticated attacker to inject malicious queries due to improper input validation and unsafe dynamic SQL handling. Successful exploitation can enable arbitrary file read/write operations and potentially lead to remote code execution.

NASM’s disasm() function contains a stack based buffer overflow when formatting disassembly output, allowing an attacker triggered out-of-bounds write when `slen` exceeds the buffer capacity.

LiteLLM through 2026-04-08 allows remote attackers to execute arbitrary code via bytecode rewriting at the /guardrails/test_custom_code URI.

Local privilege escalation due to improper handling of environment variables. The following products are affected: Acronis True Image OEM (macOS) before build 42571, Acronis True Image (macOS) before build 42902.

Improper Control of Filename for Include/Require Statement in PHP Program ('PHP Remote File Inclusion') vulnerability in Case Themes Case Theme User case-theme-user allows PHP Local File Inclusion.This issue affects Case Theme User: from n/a through < 1.0.4.

Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting') vulnerability in Zootemplate Cerato cerato allows Reflected XSS.This issue affects Cerato: from n/a through <= 2.2.18.

Improper Control of Filename for Include/Require Statement in PHP Program ('PHP Remote File Inclusion') vulnerability in CactusThemes VideoPro videopro allows PHP Local File Inclusion.This issue affects VideoPro: from n/a through <= 2.3.8.1.

Denial of Service via Out of Memory vulnerability in Apache ActiveMQ Client, Apache ActiveMQ Broker, Apache ActiveMQ. ActiveMQ NIO SSL transports do not correctly handle TLSv1.3 handshake KeyUpdates triggered by clients. This makes it possible for a client to rapidly trigger updates which causes the broker to exhaust all its memory in the SSL engine leading to DoS. Note: TLS versions before TLSv1.3 (such as TLSv1.2) are broken but are not vulnerable to OOM. Previous TLS versions require a full handshake renegotiation which causes a connection to hang but not OOM. This is fixed as well. This issue affects Apache ActiveMQ Client: before 5.19.4, from 6.0.0 before 6.2.4; Apache ActiveMQ Broker: before 5.19.4, from 6.0.0 before 6.2.4; Apache ActiveMQ: before 5.19.4, from 6.0.0 before 6.2.4. Users are recommended to upgrade to version 6.2.4 or 5.19.5, which fixes the issue.

The Gravity SMTP plugin for WordPress is vulnerable to Missing Authorization in versions up to, and including, 2.1.4. This is due to the plugin not properly verifying that a user is authorized to perform an action. This makes it possible for authenticated attackers, with subscriber-level access and above, to uninstall and deactivate the plugin and delete plugin options. NOTE: This vulnerability is also exploitable via a Cross-Site Request Forgery vector.