CVE Catalog

A vulnerability exists in Nextlevelbuilder GoClaw versions up to 3.13.3-beta.3, specifically within the exec approval management. The issue arises because the 'matchesAllowlist' function trusts basename-only entries, allowing operators to manipulate and execute commands that bypass the intended approval process. This vulnerability can be exploited remotely by authenticated operators via the 'POST /v1/tools/invoke' HTTP endpoint.

A vulnerability exists in GoClaw versions up to 3.13.2, specifically within the WebSocket approval endpoint. The issue arises in the 'RequestApproval' function of 'internal/tools/exec_approval.go', where the approval cache incorrectly handles 'allow-always' decisions. This flaw allows an operator to approve one BusyBox-wrapped command, which can then be exploited to execute different commands without additional approval, bypassing GoClaw's intended authorization process for executing host commands.

A vulnerability exists in @fastify/http-proxy versions through 11.5.0, where the request prefix is not properly rewritten if the prefix segment contains URL-encoded characters. Fastify's routing system decodes URLs for matching purposes, but the original encoded request.url is forwarded to the upstream server without modification. This allows an attacker to access upstream paths that should be hidden by the prefix rewrite, potentially exposing internal or administrative endpoints. The issue arises because the prefix rewrite process uses a literal string replacement, which fails to account for URL encoding.

A heap-based buffer overflow vulnerability has been identified in OpenPLC Runtime v3, specifically in the Modbus Master component. The issue arises in the getData() function within webserver/core/modbus_master.cpp, where the function reads data into a user-supplied buffer without proper size validation or bounds checking. This vulnerability affects all versions of OpenPLC Runtime v3 up to the latest master commit (b470206). An authenticated attacker with access to the OpenPLC web interface can exploit this vulnerability by sending a crafted HTTP POST request to the /modbus endpoint, using an oversized device_name value. The overflow occurs as the value is saved to mbconfig.cfg and parsed upon loading, overwriting adjacent struct fields and causing heap corruption. This heap corruption leads to a runtime crash, disrupting the PLC process control loop and causing a denial-of-service condition.

A server-side request forgery (SSRF) vulnerability has been identified in SurrealDB versions prior to 2.2.2. The issue arises because the database fails to properly validate HTTP redirects in its HTTP functions. This flaw enables authenticated users to circumvent deny-net restrictions by redirecting to blocked IP addresses. Attackers can exploit this by hosting a public server that redirects to denied network targets, thereby accessing internal endpoints and potentially retrieving sensitive information.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 2.0.5, 2.1.x prior to 2.1.5, and 2.2.x prior to 2.2.2. This vulnerability allows authenticated users with OWNER or EDITOR permissions at the root, namespace, or database level to create custom database functions using the DEFINE FUNCTION statement. The issue arises from the ability to nest FOR loops, each capable of iterating up to 1,000,000 times. While a single loop's iterations are limited, nesting multiple loops can lead to functions that consume excessive CPU resources. This unchecked execution monopolizes server processing, causing the database to become unresponsive to other queries and connections until it is manually restarted.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 2.0.5, 2.1.x prior to 2.1.5, and 2.2.x prior to 2.2.2. The issue arises because these versions do not impose a default execution-time limit on embedded JavaScript functions when scripting is enabled. This lack of restriction allows authenticated attackers to submit long-running scripts that can exhaust server resources, leading to a denial-of-service condition. Scripting is disabled by default, but can be enabled with specific command-line flags or environment variables.

A memory exhaustion vulnerability has been identified in SurrealDB versions prior to 2.2.2. This issue arises in the string::replace function, which fails to limit the length of resulting strings when regex patterns are used. An authenticated attacker can exploit this by crafting a malicious query that leads to unbounded string allocations, exhausting server memory and causing a denial-of-service condition.

A local file read vulnerability has been identified in SurrealDB versions prior to 2.2.2. This issue allows authenticated users with root, namespace, or database level privileges to use the DEFINE ANALYZER statement to access arbitrary files on the file system. The vulnerability specifically targets two-column tab-separated values (TSV) files, which can be read and exfiltrated by the attacker.

A vulnerability exists in SurrealDB versions prior to 2.2.2, 2.1.5, and 2.0.5, when scripting is enabled. The issue arises because the database fails to properly enforce recursion limits. Authenticated attackers can exploit this by chaining native functions with embedded JavaScript that issues new queries, creating infinite recursion that exhausts server memory. This vulnerability only affects SurrealDB instances with scripting capabilities enabled, either through command-line flags or environment variables.

A SurrealQL injection vulnerability has been identified in SurrealDB versions prior to 2.0.5, 2.1.5, and 2.2.2. The issue arises because the command-line export feature does not properly sanitize table and field names. An authenticated System User with OWNER or EDITOR roles can create tables or fields with malicious names that include SurrealQL. When a higher-privileged user imports the exported backup, the injected SurrealQL is executed, leading to privilege escalation and a complete takeover of the SurrealDB instance. Additionally, applications that allow users to define custom tables or fields are vulnerable to a second-order SurrealQL injection, even if query parameters are sanitized.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 2.2.2. This issue arises from an uncaught exception in the net module, allowing authenticated users to crash the database. Attackers can exploit this vulnerability by sending crafted HTTP queries with null bytes to the /sql endpoint. The unhandled exception causes the SurrealDB instance to crash, along with any dependent applications.

A vulnerability exists in SurrealDB versions prior to 2.2.6, 2.3.6, 2.1.8, and 3.0.0-alpha.7 that allows authenticated users to bypass network access restrictions. The issue arises because the application fails to validate DNS-resolved hostnames against deny-net restrictions in its HTTP functions. This flaw enables access to restricted internal endpoints, potentially allowing the retrieval or alteration of sensitive information and credentials, depending on the deployment scenario.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 1.1.0. The issue arises because the SurrealQL parser does not properly enforce recursion depth limits when processing nested statements and idioms, such as 'IF', 'RELATE', and attribute access. This flaw allows authorized attackers to submit queries with excessive nesting, leading to a stack overflow that crashes the server.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 1.1.1. The issue arises because the database fails to properly validate the invocation of custom parameters and functions at the root or namespace levels. Authorized clients can exploit this flaw by invoking these entities at unsupported levels, leading to a server panic and causing the SurrealDB server to crash.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 1.1.0. The issue arises in the HTTP REST API when the ID, DB, and NS headers contain special characters. Unauthenticated attackers can exploit this vulnerability by sending crafted HTTP requests with malformed header values, leading to an uncaught exception that crashes the server.

A vulnerability exists in SurrealDB versions prior to 2.0.4, where the database fails to properly enforce field permissions during SELECT, UPDATE, and DELETE operations. This oversight allows authorized users to access unauthorized field values through various query techniques. Exploitation can occur via SELECT VALUE operations, field aliasing, function arguments, WHERE clause filtering, RETURN BEFORE clauses, and SET clause references, enabling the leakage of protected field contents despite the absence of SELECT permissions.

A format string vulnerability has been identified in SurrealDB versions prior to 1.1.1. This issue resides in the rquickjs Exception::throw_type function, which is used to execute scripting functions within SurrealDB. When scripting is enabled, attackers with scripting privileges can inject format string sequences into error inputs, potentially leading to arbitrary memory read or execution of code with SurrealDB process privileges.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 1.2.0. This issue arises from an uncaught exception in the query executor, which occurs when the server processes calls to nonexistent built-in functions. Authorized clients can exploit this vulnerability by crafting pre-parsed queries that invoke these nonexistent functions, leading to a server panic and subsequent crash.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 1.2.1. This issue arises from uncaught exception handling in the span rendering process when queries contain errors related to line terminator characters. Authorized clients can exploit this vulnerability by submitting malformed queries that cause a panic in the span rendering code, leading to a server crash.

A vulnerability exists in SurrealDB versions prior to 1.5.4 and 2.0.0-alpha.6, where the database authentication process is not properly validated when a scope user switches databases using the 'USE' clause or method. This flaw allows attackers with an authenticated session to impersonate an unrelated user in a different database, provided a user record with the same identifier exists. Such impersonation could enable unauthorized actions if the user's permissions are solely based on the $auth parameter.

A vulnerability exists in SurrealDB versions prior to 1.5.5 and 2.0.0-beta prior to 2.0.0-beta.3, where the RPC API's sign-in and sign-up operations accept arbitrary objects without proper validation for non-computed values. This flaw allows an unauthenticated attacker to encode a binary object containing a subquery using the bincode serialization format and replace the credentials with it. The injected subquery is executed within the context of the database owner's sign-in or sign-up query, under a system user session with editor privileges. As a result, the attacker can manipulate non-IAM resources by selecting, creating, updating, or deleting them, although they cannot directly view the query results or affect IAM resources, which require owner privileges.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 2.0.4. The issue arises from uncaught exceptions in the parser's error handling when processing empty strings. Authorized clients can exploit this by sending malformed queries that convert empty strings into record, duration, or datetime types. This conversion failure triggers a panic in the error rendering process, causing the server to crash.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 2.1.0. The issue arises in the sorting mechanism when the ORDER BY rand() clause is used. Authorized clients can execute queries with this clause, triggering a panic in the sorting function that crashes the server.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 2.1.0. The issue arises in the role conversion process, where privileged owner users can assign nonexistent roles to users. When a user with an invalid role attempts to sign in, it triggers an uncaught panic, causing the server to crash.

A denial-of-service vulnerability has been identified in SurrealDB versions prior to 2.1.0. The issue arises from an uncaught exception in the rand::time() function, which generates a random time from an optional range of Unix timestamps. The function can return None, leading to a panic when unwrap is called, causing server crashes. Authorized clients can exploit this by repeatedly invoking rand::time(), reliably triggering server panics and causing a denial-of-service condition.

A vulnerability exists in SurrealDB versions prior to 2.1.4, where the 'DEFINE TABLE ... OVERWRITE' clause fails to properly update table definitions for tables designated as 'TYPE RELATION'. This issue arises because the 'OVERWRITE' clause does not effectively modify the 'PERMISSIONS' settings for these tables. Consequently, an administrator may mistakenly believe that permission changes have been applied, while in reality, they have not. This oversight can allow clients with query authorization to access data in the table that the intended permission updates were meant to restrict.

A vulnerability exists in SurrealDB versions prior to 1.0.1, where default table permissions are set to FULL instead of NONE. This misconfiguration allows SELECT, CREATE, UPDATE, and DELETE operations on tables without explicit permissions. As a result, attackers with database access or unauthenticated users on publicly exposed instances can perform unrestricted operations on unprotected tables within their authorization scope.

A code injection vulnerability exists in Uproot versions through 5.7.4. This issue arises because Uproot generates Python class source code from ROOT TStreamerInfo records and compiles it at runtime. Some metadata fields, such as streamer element names, are interpolated into the generated Python code without proper escaping. An attacker can exploit this by crafting a ROOT file that includes disruptive Python expressions in the metadata. When Uproot processes the file and calls the corresponding reader method, the injected code is executed in the application's context, leading to arbitrary code execution.

A denial-of-service vulnerability has been identified in Apache Traffic Server, specifically in the HTTP/2 protocol, where stalled flow-control can be exploited. This issue affects Apache Traffic Server versions 9.0.0 prior to 9.1.14 and 10.0.0 prior to 10.1.3.

A vulnerability in @fastify/reply-from versions 8.3.1 prior to 12.6.4 allows for cross-upstream request routing due to a URL cache key collision. The vulnerability arises because the internal URL cache key is created by concatenating the destination and source paths without a delimiter. This can lead to different destination-source pairs producing the same cache key while pointing to different upstream URLs. As a result, a cached URL intended for one upstream can be incorrectly reused for a request meant for another, causing unauthorized access to and modification of data across upstreams. The default configuration is vulnerable, but setting disableCache to true can prevent this issue.

A vulnerability exists in @fastify/http-proxy versions 9.4.0 prior to 11.5.0, allowing WebSocket requests to escape the intended rewrite prefix and access upstream endpoints that should be hidden. This is achieved by exploiting the WebSocket routing path, which normalizes dot segments and can be manipulated to traverse out of the designated prefix. The issue arises because the proxy does not properly validate WebSocket destination paths against the rewrite prefix, creating a path traversal vulnerability. Exploitation requires a non-normalizing WebSocket client, such as a raw HTTP client or one that forwards requests without modification.

A stack-based buffer overflow vulnerability has been identified in Shibby Tomato version 1.28, specifically within the Scheduler Name Handler component. The issue arises in the function 'sub_42537C', where the manipulation of the 'a1' argument leads to a buffer overflow. This vulnerability can be exploited remotely, allowing for potential command injection via the 'system()' function.

A stack-based buffer overflow vulnerability has been identified in Shibby Tomato version 1.28 RT-N5x MIPSR2 Build 124. The issue arises in the web server component, specifically within the web monitor's log rendering function. This vulnerability can be exploited remotely.

A stack out-of-bounds write vulnerability has been identified in Shibby Tomato version 1.28 RT-N5x MIPSR2 Build 124. The issue arises in the `setup_conntrack` function within the `sbin/rc` file, where the `ct_tcp_timeout` argument is manipulated. This flaw allows for an out-of-bounds write by improperly handling TCP timeout values, which can be exploited remotely.

A path traversal vulnerability has been identified in the Halo Files Backup Endpoint component, specifically in versions through 2.24.2. The issue arises in the Download function of MigrationEndpoint.java, where the status.filename field is not properly validated. This flaw allows remote attackers to manipulate the filename parameter, potentially leading to the unauthorized reading of sensitive files from the server.

A vulnerability in Sipeed PicoClaw versions through 0.2.9 allows for the inclusion of untrusted repository-local skills metadata into the AI model's system prompt. This issue arises because PicoClaw automatically loads skill information from the current working directory when the 'PICOCLAW_BUILTIN_SKILLS' environment variable is not set. The vulnerability can be exploited locally by placing a malicious 'SKILL.md' file in a cloned repository's 'skills' directory, which is then injected into the model's prompt without any explicit trust or approval. As a result, an attacker could manipulate the AI's behavior by controlling the prompt content before the user has a chance to review or authorize it.

A directory traversal vulnerability has been identified in VMware Avi Load Balancer. This issue arises from inadequate file path validation, allowing malicious authenticated network users to perform directory traversal attacks. The vulnerability is present in versions 31.1.1 through 31.2.2, 30.1.1 through 30.2.6, and 22.1.1 through 22.1.7, with version 32.1.1 being the most recent vulnerable release.

A privilege escalation vulnerability has been identified in VMware Avi Load Balancer, allowing a malicious authenticated user with network access to execute remote code. This issue is present in versions 31.1.1 through 31.2.2, 30.1.1 through 30.2.6, and 22.1.1 through 22.1.7, with version 32.1.1 being the most recent vulnerable release.

A remote code execution vulnerability exists in VMware Avi Load Balancer. This issue allows a malicious authenticated user with network access to inject and execute code. The vulnerability is present in versions 31.1.1 through 31.2.2, 30.1.1 through 30.2.6, and 22.1.1 through 22.1.7, with version 32.1.1 being the most recent vulnerable release.

A local privilege escalation vulnerability has been identified in VMware Avi Load Balancer. This issue allows a malicious user with local access to escalate privileges and execute code as the root user. The vulnerability is present in versions 31.1.1 through 31.2.2, 30.1.1 through 30.2.6, and 22.1.1 through 22.1.7, with version 32.1.1 being the only affected release in the 32.x series.

A remote code execution vulnerability has been identified in VMware Avi Load Balancer. This issue allows a malicious user with network access to the Avi Control Plane to execute code remotely. The vulnerability is present in versions 31.1.1 through 31.2.2, 30.1.1 through 30.2.6, and 22.1.1 through 22.1.7, with version 32.1.1 being the only exception. The vulnerability arises from flaws in the application's handling of network requests, which can be exploited to inject and execute arbitrary code on the server.

An authorization bypass vulnerability has been identified in VMware Avi Load Balancer. This vulnerability allows a malicious actor on the network to access a limited subset of the Avi Control Plane without proper authorization. The issue is present in VMware Avi Load Balancer versions 31.1.1 through 31.2.2, 30.1.1 through 30.2.6, and 22.1.1 through 22.1.7, with version 32.1.1 being the only version currently available that is not vulnerable.

A critical authentication bypass vulnerability has been identified in VMware Avi Load Balancer versions 31.1.1 through 31.2.2, 30.1.1 through 30.2.6, and 22.1.1 through 22.1.7. This vulnerability allows a malicious user with network access to bypass authentication and access the Avi Control plane. The issue has been evaluated with a CVSSv3 base score of 9.8, indicating its critical severity.

A server-side request forgery (SSRF) vulnerability has been identified in Sipeed PicoClaw versions through 0.2.9. The issue arises in the web_fetch function within the file pkg/tools/integration/web.go. This vulnerability allows remote exploitation by bypassing the application's IP validation checks, particularly for ISATAP IPv6 addresses that embed private IPv4 loops. As a result, crafted URLs can be used to access internal services that should be protected.

An authentication bypass vulnerability has been identified in Sipeed PicoClaw versions through 0.2.9, specifically within the LINE webhook component. The issue arises in the 'webhook.ParseRequest' function of 'pkg/channels/line/line.go', where the webhook handler validates the HMAC signature but fails to suppress duplicate events. This flaw allows an attacker to replay previously valid LINE webhooks, causing the same event to be processed multiple times. The vulnerability can be exploited remotely, leading to unintended duplicate actions within the application.

A time-of-check time-of-use (TOCTOU) vulnerability has been identified in Sipeed PicoClaw versions through 0.2.9. The issue arises in the ExecTool.executeRun function within the file pkg/agent/pipeline_execute.go. This vulnerability allows an attacker to manipulate the 'cwd' argument, creating a race condition that disrupts the approval process for command execution. The exploitation must be performed locally, and the available public exploit could be used to trigger this vulnerability.

A cross-site request forgery (CSRF) vulnerability has been identified in Sipeed PicoClaw versions through 0.2.9. The issue arises in the launcher authentication setup endpoint, which is publicly accessible during the initial setup phase. The endpoint lacks proper validation of origin headers, allowing remote attackers to manipulate the setup process and gain control over the user's PicoClaw launcher.

A vulnerability exists in AstrBot versions through 4.25.5, specifically within the Filesystem Computer-Use Tool. The issue arises in the function _normalize_rw_path, located in astrbot/core/tools/computer_tools/fs.py. This vulnerability allows local users to bypass workspace restrictions by exploiting hardlink aliases, enabling unauthorized access to files outside the designated workspace boundary. The AstrBot process must have the appropriate permissions to the target files for this exploit to be successful.

An authentication bypass vulnerability has been identified in AstrBot versions through 4.25.5. The issue arises in the OpenAPI chat functionality, where the 'username' parameter can be manipulated to spoof identities. This vulnerability allows holders of 'chat'-scope API keys to impersonate administrators and execute commands reserved for them. The flaw exists in the API component, specifically within the 'OpenApiRoute.chat_send' function of 'astrbot/dashboard/routes/open_api.py'. The vulnerability can be exploited remotely, and has been publicly disclosed.