ghsa

### Summary The Fides Webserver API's built-in IP-based rate limiting is ineffective in environments with CDNs, proxies or load balancers. The system incorrectly applies rate limits based on directly connected infrastructure IPs rather than client IPs, and stores counters in-memory rather than in a shared store. This allows attackers to bypass intended rate limits and potentially cause denial of service. This vulnerability only affects deployments relying on Fides's built-in rate limiting for protection. Deployments using external rate limiting solutions (WAFs, API gateways, etc.) are not affected. ### Details The vulnerability has two components: 1. Rate limiting uses the immediate connection source IP instead of the actual client IP 2. Rate limit counters are maintained in-memory per container rather than in a shared store In production environments, these issues allow clients to exceed intended rate limits and enable attackers to trigger rate limits on infrastructure IPs, caus...

### Summary The Fides Admin UI login endpoint relies on a general IP-based rate limit for all API traffic and lacks specific anti-automation controls designed to protect against brute-force attacks. This could allow attackers to conduct credential testing attacks, such as credential stuffing or password spraying, which poses a risk to accounts with weak or previously compromised passwords. ### Details Fides uses a configurable, system-wide rate limit to control traffic from any single IP address. Because this single limit must be set high enough to accommodate endpoints that receive a large volume of legitimate traffic, it offers only weak protection for the login endpoint. The system is not equipped with more advanced protections tailored specifically for authentication ### Impact Although password complexity requirements and the global rate limit make a traditional brute-force attack against a single account difficult, the lack of authentication-specific protections exposes Fid...

### Summary Admin UI user password changes in Fides do not invalidate active user sessions, creating a vulnerability chaining opportunity where attackers who have obtained session tokens through other attack vectors (such as XSS) can maintain access even after password reset. This issue is not directly exploitable on its own and requires a prerequisite vulnerability to obtain valid session tokens in the first place. ### Details Fides uses encrypted authentication tokens with extended expiration periods. When a password is changed via password reset endpoints, the system updates the password hash in the database but does not invalidate existing client sessions or tokens. The authentication system validates tokens based on their cryptographic integrity and expiration time, not against the current password state. The frontend application stores authentication state in browser local storage, which persists across browser sessions until explicit logout or natural token expiration. This...

# Command Injection in MCP Server The MCP Server at https://github.com/akoskm/create-mcp-server-stdio is written in a way that is vulnerable to command injection vulnerability attacks as part of some of its MCP Server tool definition and implementation. ## Vulnerable tool The MCP Server exposes the tool `which-app-on-port` which relies on Node.js child process API `exec` which is an unsafe and vulnerable API if concatenated with untrusted user input. Vulnerable line of code: https://github.com/akoskm/create-mcp-server-stdio/blob/main/src/index.ts#L24-L40 ```js server.tool("which-app-on-port", { port: z.number() }, async ({ port }) => { const result = await new Promise<ProcessInfo>((resolve, reject) => { exec(`lsof -t -i tcp:${port}`, (error, pidStdout) => { if (error) { reject(error); return; } const pid = pidStdout.trim(); exec(`ps -p ${pid} -o comm=`, (error, stdout) => { if (error) { reject(error); return...

CodeceptJS 3.7.3 contains a command injection vulnerability in the emptyFolder function (lib/utils.js). The execSync command directly concatenates the user-controlled directoryPath parameter without sanitization or escaping, allowing attackers to execute arbitrary commands.

An arbitrary file upload vulnerability in the Chat Trigger component of N8N v1.95.3, v1.100.1, and v1.101.1 allows attackers to execute arbitrary code via uploading a crafted HTML file.

An issue was discovered in Django 4.2 before 4.2.24, 5.1 before 5.1.12, and 5.2 before 5.2.6. FilteredRelation is subject to SQL injection in column aliases, using a suitably crafted dictionary, with dictionary expansion, as the **kwargs passed QuerySet.annotate() or QuerySet.alias().

There is a serialization of Untrusted Data vulnerability in Apache Jackrabbit Core and Apache Jackrabbit JCR Commons. This issue affects Apache Jackrabbit Core: from 1.0.0 through 2.22.1; Apache Jackrabbit JCR Commons: from 1.0.0 through 2.22.1. Deployments that accept JNDI URIs for JCR lookup from untrusted users allows them to inject malicious JNDI references, potentially leading to arbitrary code execution through deserialization of untrusted data. Users are recommended to upgrade to version 2.22.2. JCR lookup through JNDI has been disabled by default in 2.22.2. Users of this feature need to enable it explicitly and are adviced to review their use of JNDI URI for JCR lookup.

### Summary Atlantis publicly exposes detailed version information on its `/status` endpoint. This information disclosure could allow attackers to identify and target known vulnerabilities associated with the specific versions, potentially compromising the service's security posture. ### Details The `/status` endpoint in Atlantis returns not only a health check but also detailed version and build information. This disclosure violates the principle of minimizing exposed sensitive metadata and can be leveraged by adversaries to correlate the version information with public vulnerability databases, including CVE listings. Although Atlantis is a public repository maintained by an external team, reducing this exposure can lessen the overall risk of targeted attacks. For example, the source code handling the `/status` endpoint exposes version details that allow one to infer software dependencies and system configurations. Best practices, including guidelines from the [OWASP Top 10](https:/...

### Summary Provided grammar, would fit in a context window of most of the models, but takes minutes to process in 0.1.23. In testing with 0.1.16 the parser worked fine so this seems to be a regression caused by Earley parser. ### Details Full reproducer provider in the POC section. The resulting grammar is around 70k tokens, and the grammar parsing itself (with the models I checked) was significantly longer than LLM processing itself, meaning this can be used to DOS model providers. ### Patch This problem is caused by the grammar optimizer introduced in v0.1.23 being too slow. It only happens for very large grammars (>100k characters), like the below one. v0.1.24 solved this problem by optimizing the speed of the grammar optimizer and disable some slow optimization for large grammars. Thanks to @Seven-Streams ### PoC ``` import string import random def enum_schema(size=10000,str_len=10): enum = {"enum": ["".join(random.choices(string.ascii_uppercase, k=str_len)) for _ in...