The OAuth implementation in workers-oauth-provider that is part of MCP framework https://github.com/cloudflare/workers-mcp , did not correctly validate that redirect_uri was on the allowed list of redirect URIs for the given client registration. Fixed in:  https://github.com/cloudflare/workers-oauth-provider/pull/26 https://github.com/cloudflare/workers-oauth-provider/pull/26 Impact: Under certain circumstances (see below), if a victim had previously authorized with a server built on workers-oath-provider, and an attacker could later trick the victim into visiting a malicious web site, then attacker could potentially steal the victim's credentials to the same OAuth server and subsequently impersonate them. In order for the attack to be possible, the OAuth server's authorized callback must be designed to auto-approve authorizations that appear to come from an OAuth client that the victim has authorized previously. The authorization flow is not implemented by workers-oauth-provider; it is up to the application built on top to decide whether to implement such automatic re-authorization. However, many applications do implement such logic. Note: It is a basic, well-known requirement that OAuth servers should verify that the redirect URI is among the allowed list for the client, both during the authorization flow and subsequently when exchanging the authorization code for an access token. workers-oauth-provider implemented only the latter check, not the former. Unfortunately, the former is the much more important check. Readers who are familiar with OAuth may recognize that failing to check redirect URIs against the allowed list is a well-known, basic mistake, covered extensively in the RFC and elsewhere. The author of this library would like everyone to know that he was, in fact, well-aware of this requirement, thought about it a lot while designing the library, and then, somehow, forgot to actually make sure the check was in the code. That is, it's not that he didn't know what he was doing, it's that he knew what he was doing but flubbed it.

All versions of the package drogonframework/drogon are vulnerable to HTTP Response Splitting when untrusted user input is used to build header values in the addHeader and addCookie functions. An attacker can add the \r\n (carriage return line feeds) characters to end the HTTP response headers and inject malicious content.

A vulnerability in the VPN web client services component of Cisco Adaptive Security Appliance (ASA) Software and Cisco Firepower Threat Defense (FTD) Software could allow an unauthenticated, remote attacker to conduct browser-based attacks against users of an affected device. This vulnerability is due to improper validation of input that is passed to the VPN web client services component before being returned to the browser that is in use. An attacker could exploit this vulnerability by persuading a user to visit a website that is designed to pass malicious requests to a device that is running Cisco ASA Software or Cisco FTD Software and has web services endpoints supporting VPN features enabled. A successful exploit could allow the attacker to reflect malicious input from the affected device to the browser that is in use and conduct browser-based attacks, including cross-site scripting attacks. The attacker could not directly impact the affected device.

This vulnerability occurs when a web server fails to correctly process the Content-Length of POST requests. This can lead to HTTP request smuggling or XSS.

laminas-diactoros is a PHP package containing implementations of the PSR-7 HTTP message interfaces and PSR-17 HTTP message factory interfaces. Applications that use Diactoros, and are either not behind a proxy, or can be accessed via untrusted proxies, can potentially have the host, protocol, and/or port of a `Laminas\Diactoros\Uri` instance associated with the incoming server request modified to reflect values from `X-Forwarded-*` headers. Such changes can potentially lead to XSS attacks (if a fully-qualified URL is used in links) and/or URL poisoning. Since the `X-Forwarded-*` headers do have valid use cases, particularly in clustered environments using a load balancer, the library offers mitigation measures only in the v2 releases, as doing otherwise would break these use cases immediately. Users of v2 releases from 2.11.1 can provide an additional argument to `Laminas\Diactoros\ServerRequestFactory::fromGlobals()` in the form of a `Laminas\Diactoros\RequestFilter\RequestFilterInterface` instance, including the shipped `Laminas\Diactoros\RequestFilter\NoOpRequestFilter` implementation which ignores the `X-Forwarded-*` headers. Starting in version 3.0, the library will reverse behavior to use the `NoOpRequestFilter` by default, and require users to opt-in to `X-Forwarded-*` header usage via a configured `Laminas\Diactoros\RequestFilter\LegacyXForwardedHeaderFilter` instance. Users are advised to upgrade to version 2.11.1 or later to resolve this issue. Users unable to upgrade may configure web servers to reject `X-Forwarded-*` headers at the web server level.

netius prior to 1.17.58 is vulnerable to HTTP Request Smuggling. HTTP pipelining issues and request smuggling attacks might be possible due to incorrect Transfer encoding header parsing which could allow for CL:TE or TE:TE attacks.

meinheld prior to 1.0.2 is vulnerable to HTTP Request Smuggling. HTTP pipelining issues and request smuggling attacks might be possible due to incorrect Content-Length and Transfer encoding header parsing.