ghsa

An SQL Injection vulnerability in Apache Superset exists due to improper neutralization of special elements used in SQL commands. Specifically, certain engine-specific functions are not checked, which allows attackers to bypass Apache Superset's SQL authorization. To mitigate this, a new configuration key named DISALLOWED_SQL_FUNCTIONS has been introduced. This key disallows the use of the following PostgreSQL functions: version, query_to_xml, inet_server_addr, and inet_client_addr. Additional functions can be added to this list for increased protection. This issue affects Apache Superset: before 4.0.2. Users are recommended to upgrade to version 4.0.2, which fixes the issue.

### Impact Editors that use `MediaEmbedElement` and pass custom `urlParsers` to the `useMediaState` hook may be vulnerable to XSS if a custom parser allows `javascript:`, `data:` or `vbscript:` URLs to be embedded. Editors that do not use `urlParsers` and instead consume the `url` property directly may also be vulnerable if the URL is not sanitised. The default parsers `parseTwitterUrl` and `parseVideoUrl` are not affected. Examples of vulnerable code: ```tsx const { embed } = useMediaState({ urlParsers: [ // Custom parser that does not use an allowlist or validate the URL protocol (url) => ({ url }), ], }); return ( <iframe src={embed!.url} // ... /> ); ``` ```tsx const { url } = useMediaState(); return ( <iframe // url property used directly from useMediaState() with no sanitisation src={url} // ... /> ); ``` ```tsx const { url } = element; return ( <iframe // url property used directly from element with no sanitisation src={...

RISC Zero zkVM was designed from its inception to provide three main guarantees: 1. *Computational integrity*: that a given software program executed correctly. 2. *Succinctness*: that the proof of execution does not grow in relation to the program being executed. 3. *Zero Knowledge*: that details of the program execution are not visible within the proof of program execution. Ulrich Habock and Al Kindi have released [new research] that indicates that several STARK implementations -including our RISC Zero zkVM- do not meet the requirements to assert the specific property of zero knowledge provably. While a vast majority of real-world applications that leverage RISC Zero zkVM or similar systems depend primarily on computational integrity and succinctness, a subset of applications critically depend on the privacy guarantees provided by zero-knowledge; and for those use cases, users are cautioned to understand the research and make informed decisions based on the risks outlined in using...

### Summary Applications using the `zitadel-go` `v3` library (`next` branch) might be impacted by package vulnerabilities. The output of `govulncheck` suggests that only `example` code seems to be impacted, based on 1 of the 3 potential vulnerabilities. This vulnerability is located in the transitive dependency `golang.org/x/net v0.19.0`, [CVE-2023-45288](https://www.cve.org/CVERecord?id=CVE-2023-45288) ### Patches 3.0.0-next versions are fixed on >= [3.0.0-next.3](https://github.com/zitadel/zitadel-go/releases/tag/v3.0.0-next.3) ZITADEL recommends upgrading to the latest versions available in due course. ### Workarounds If updating the zitadel-go library is not an option, updating the affected (transient) dependencies works as a workaround. ### Details #### Direct deps: - [GO-2024-2631](https://pkg.go.dev/vuln/GO-2024-2631) Decompression bomb vulnerability in github.com/go-jose/go-jose - github.com/go-jose/go-jose/v3 Fixed in v3.0.3. This module is necessary because [github....

### Summary HTTP `OPTIONS` requests are always allowed by `OpaMiddleware`, even when they lack authentication, and are passed through directly to the application. The maintainer uncertain whether this should be classed as a "bug" or "security issue" – but is erring on the side of "security issue" as an application could reasonably assume OPA controls apply to *all* HTTP methods, and it bypasses more sophisticated policies. ### Details `OpaMiddleware` allows all HTTP `OPTIONS` requests without evaluating it against any policy: https://github.com/busykoala/fastapi-opa/blob/6dd6f8c87e908fe080784a74707f016f1422b58a/fastapi_opa/opa/opa_middleware.py#L79-L80 If an application provides different responses to HTTP `OPTIONS` requests based on an entity existing (such as to indicate whether an entity is writable on a system level), an unauthenticated attacker could discover which entities exist within an application (CWE-204). ### PoC This toy application is based on the behaviour of an ...

### Summary In `torrentpier/library/includes/functions.php`, `get_tracks()` uses the unsafe native PHP serialization format to deserialize user-controlled cookies: https://github.com/torrentpier/torrentpier/blob/84f6c9f4a081d9ffff4c233098758280304bf50f/library/includes/functions.php#L41-L60 ### PoC One can use [`phpggc`](https://github.com/ambionics/phpggc/) and the chain `Guzzle/FW1` to write PHP code to an arbitrary file, and execute commands on the system. For instance, the cookie `bb_t` will be deserialized when browsing to `viewforum.php`.

### Summary The maintainer been contemplating whether FTP or other protocols could serve as useful functionalities, but there may not be a practical reason for it since we are utilizing headless Chrome to capture screenshots. The argument is based on the assumption that this package can function as a service. The package includes an `ALLOW_LIST` where the host can specify which services the user is permitted to capture screenshots of. By default, capturing screenshots of web services running on localhost, 127.0.0.1, or the [::] is allowed. The maintainer is of the opinion that the package should also have a blacklist due to a potential vulnerability (or rather design oversight). If someone hosts this on a server, users could then capture screenshots of other web services running locally. Unless this is strictly for web pages. Something similar here: https://github.com/follow-redirects/follow-redirects/issues/235 (localhost is intended for end users or hosts to deny, and the package...

### Summary When trying to add a `BLOCK_LIST` feature when the maintainer noticed they didn't sanitize the `ImageId` in the code, which leads to path traversal vulnerability. Now, this is different from a traditional path traversal issue, because as of NOW you can store the image in any place arbitrarily, and given enough time they might be able to come up with a working exploit BUT for the time being they am reporting this. ### Details @jmondi/url-to-png does not sanitizing the `ImageID` as in not removing special chars from the params [(extract_query_params.ts#l75)](https://github.com/jasonraimondi/url-to-png/blob/e43098e0af3a380ebc044e7f303a83933b94b434/src/middlewares/extract_query_params.ts#L75) ```js const imageId = dateString + "." + slugify(validData.url) +configToString(params); ``` This when fed to other parts of the code such as ([filesystem.ts#L34](https://github.com/jasonraimondi/url-to-png/blob/8afc00247c1d7e6c7b37356a5f6282b486e596fa/src/lib/storage/filesystem.ts#L...

### Summary The ProfileBasedRequestOptionsBuilder method returns allowedCredentials without any credentials if no username was found. ### Details When WebAuthn is used as the first or only authentication method, an attacker can enumerate usernames based on the absence of the `allowedCredentials` property in the assertion options response. This allows enumeration of valid or invalid usernames. #### Proposal how to resolve it: ``` return $this->publicKeyCredentialRequestOptionsFactory->create( $this->profile, count($allowedCredentials) <= 0 ? self::getRandomCredentials(): $allowedCredentials, $optionsRequest->userVerification, $extensions ); private static function getRandomCredentials(): array { $credentialSources = []; for ($i = 0; $i <= rand(0,1); $i++) { $credentialSources[] = new PublicKeyCredentialSource( random_bytes(32), "public-key", [], ...

In Apache Linkis <=1.5.0, due to the lack of effective filteringof parameters, an attacker configuring malicious `db2` parameters in the DataSource Manager Module will result in jndi injection. Therefore, the parameters in the DB2 URL should be blacklisted.  This attack requires the attacker to obtain an authorized account from Linkis before it can be carried out. Versions of Apache Linkis <=1.5.0 will be affected. We recommend users upgrade the version of Linkis to version 1.6.0.