A vulnerability in the cp utility of uutils coreutils affects recursive copy operations (-R). When copying character and block device nodes, the utility incorrectly reads device nodes as regular files instead of preserving their device semantics using mknod. This causes device nodes like /dev/null to become regular files, potentially leading to denial of service through disk exhaustion or process hangs when accessing these files.
fast-jwt provides fast JSON Web Token (JWT) implementation. From 0.0.1 to before 6.2.0, setting up a custom cacheKeyBuilder method which does not properly create unique keys for different tokens can lead to cache collisions. This could cause tokens to be mis-identified during the verification process leading to valid tokens returning claims from different valid tokens and users being mis-identified as other users based on the wrong token. Version 6.2.0 contains a patch.
A vulnerability in h3js srvx versions prior to 0.11.13 allows middleware bypass due to a pathname parsing discrepancy in the FastURL constructor. This occurs when a raw HTTP request uses an absolute URI with a non-standard scheme such as file://. The issue is resolved starting in version 0.11.13 by changing FastURL to use the native URL constructor for strings not starting with '/'. The CVSS score is 4.8, indicating medium severity.
H3 is a minimal H(TTP) framework. In versions 2.0.0-0 through 2.0.1-rc.16, the `mount()` method in h3 uses a simple `startsWith()` check to determine whether incoming requests fall under a mounted sub-application's path prefix. Because this check does not verify a path segment boundary (i.e., that the next character after the base is `/` or end-of-string), middleware registered on a mount like `/admin` will also execute for unrelated routes such as `/admin-public`, `/administrator`, or `/adminstuff`. This allows an attacker to trigger context-setting middleware on paths it was never intended to cover, potentially polluting request context with unintended privilege flags. Version 2.0.2-rc.17 contains a patch.
WeKnora is an LLM-powered framework designed for deep document understanding and semantic retrieval. Prior to version 0.3.0, a vulnerability involving tool name collision and indirect prompt injection allows a malicious remote MCP server to hijack tool execution. By exploiting an ambiguous naming convention in the MCP client (mcp_{service}_{tool}), an attacker can register a malicious tool that overwrites a legitimate one (e.g., tavily_extract). This enables the attacker to redirect LLM execution flow, exfiltrate system prompts, context, and potentially execute other tools with the user's privileges. This issue has been patched in version 0.3.0.
A Server-Side Request Forgery (SSRF) vulnerability was identified in the @opennextjs/cloudflare package, resulting from a path normalization bypass in the /cdn-cgi/image/ handler.The @opennextjs/cloudflare worker template includes a /cdn-cgi/image/ handler intended for development use only. In production, Cloudflare's edge intercepts /cdn-cgi/image/ requests before they reach the Worker. However, by substituting a backslash for a forward slash (/cdn-cgi\image/ instead of /cdn-cgi/image/), an attacker can bypass edge interception and have the request reach the Worker directly. The JavaScript URL class then normalizes the backslash to a forward slash, causing the request to match the handler and trigger an unvalidated fetch of arbitrary remote URLs.
For example:
https://victim-site.com/cdn-cgi\image/aaaa/https://attacker.com
In this example, attacker-controlled content from attacker.com is served through the victim site's domain (victim-site.com), violating the same-origin policy and potentially misleading users or other services.
Note: This bypass only works via HTTP clients that preserve backslashes in paths (e.g., curl --path-as-is). Browsers normalize backslashes to forward slashes before sending requests.
Additionally, Cloudflare Workers with Assets and Cloudflare Pages suffer from a similar vulnerability. Assets stored under /cdn-cgi/ paths are not publicly accessible under normal conditions. However, using the same backslash bypass (/cdn-cgi\... instead of /cdn-cgi/...), these assets become publicly accessible. This could be used to retrieve private data. For example, Open Next projects store incremental cache data under /cdn-cgi/_next_cache, which could be exposed via this bypass.
SmarterTools SmarterMail versions prior to build 9518 contain an unauthenticated path coercion vulnerability in the background-of-the-day preview endpoint. The application base64-decodes attacker-supplied input and uses it as a filesystem path without validation. On Windows systems, this allows UNC paths to be resolved, causing the SmarterMail service to initiate outbound SMB authentication attempts to attacker-controlled hosts. This can be abused for credential coercion, NTLM relay attacks, and unauthorized network authentication.
CVE-2025-64750 is a medium severity vulnerability in Sylabs Singularity container platforms (SingularityCE and SingularityPRO) prior to versions 4.3.5 and 4.1.11. It involves a symbolic link (symlink) following issue where an attacker can bypass Linux Security Module (LSM) restrictions by redirecting the mount of /proc to a shared mount controlled by the attacker. Exploitation requires the victim to run a malicious container image that mounts /proc to this attacker-controlled location. The attacker must also have control over the shared mount content, either via another malicious container or host permissions. This vulnerability can lead to partial compromise of confidentiality, integrity, and availability of the container environment. No known exploits are reported in the wild.
When zx is invoked with --prefer-local=<path>, the CLI creates a symlink named ./node_modules pointing to <path>/node_modules. Due to a logic error in src/cli.ts (linkNodeModules / cleanup), the function returns the target path instead of the alias (symlink path). The later cleanup routine removes what it received, which deletes the target directory itself. Result: zx can delete an external <path>/node_modules outside the current working directory.
CommandKit is the discord.js meta-framework for building Discord bots. In versions 1.2.0-rc.1 through 1.2.0-rc.11, a logic flaw exists in the message command handler that affects how the commandName property is exposed to both middleware functions and command execution contexts when handling command aliases. When a message command is invoked using an alias, the ctx.commandName value reflects the alias rather than the canonical command name. This occurs in both middleware functions and within the command's own run function. Although not explicitly documented, CommandKit's examples and guidance around middleware usage implicitly convey that ctx.commandName represents the canonical command identifier. Middleware examples in the documentation consistently use ctx.commandName to reference the command being executed. Developers who assume ctx.commandName is canonical may introduce unintended behavior when relying on it for logic such as permission checks, rate limiting, or audit logging. This could allow unauthorized command execution or inaccurate access control decisions. Slash commands and context menu commands are not affected. This issue has been patched in version 1.2.0-rc.12, where ctx.commandName now consistently returns the actual canonical command name regardless of the alias used to invoke it.
Hono is a Web application framework that provides support for any JavaScript runtime. Versions 4.8.0 through 4.9.5 contain a flaw in the getPath utility function which could allow path confusion and potential bypass of proxy-level ACLs (e.g. Nginx location blocks). The original implementation relied on fixed character offsets when parsing request URLs. Under certain malformed absolute-form Request-URIs, this could lead to incorrect path extraction depending on the application and environment. If proxy ACLs are used to protect sensitive endpoints such as /admin, this flaw could have allowed unauthorized access. The confidentiality impact depends on what data is exposed: if sensitive administrative data is exposed, the impact may be high, otherwise it may be moderate. This issue is fixed in version 4.9.6.