Threats affecting pkg:composer/saloonphp/saloon

CVE-2026-33942 is a high-severity vulnerability in the saloonphp saloon PHP library versions prior to 4.0.0. It involves unsafe deserialization of untrusted data in the AccessTokenAuthenticator::unserialize() method, which uses PHP's unserialize() with allowed_classes set to true. An attacker controlling the serialized input can inject malicious objects that trigger PHP magic methods, potentially leading to remote code execution when combined with common dependencies like Monolog. The vulnerability is fixed in version 4.0.0 by removing PHP serialization from the affected class.

Saloon is a PHP library that gives users tools to build API integrations and SDKs. Prior to version 4.0.0, fixture names were used to build file paths under the configured fixture directory without validation. A name containing path segments (e.g. ../traversal or ../../etc/passwd) resulted in a path outside that directory. When the application read a fixture (e.g.for mocking) or wrote one (e.g.when recording responses), it could read or write files anywhere the process had access. If the fixture name was derived from user or attacker-controlled input (e.g.request parameters or config), this constituted a path traversal vulnerability and could lead to disclosure of sensitive files or overwriting of critical files. The fix in version 4.0.0 adds validation in the fixture layer (rejecting names with /, \, .., or null bytes, and restricting to a safe character set) and defense-in-depth in the storage layer (ensuring the resolved path remains under the base directory before any read or write).

Saloon is a PHP library that gives users tools to build API integrations and SDKs. Prior to version 4.0.0, when building the request URL, Saloon combined the connector's base URL with the request endpoint. If the endpoint was a valid absolute URL, the code used that URL as-is and ignored the base URL. The request—and any authentication headers, cookies, or tokens attached by the connector—was then sent to the attacker-controlled host. If the endpoint could be influenced by user input or configuration (e.g.redirect_uri, callback URL), this allowed server-side request forgery (SSRF) and/or credential leakage to a third-party host. The fix in version 4.0.0 is to reject absolute URLs in the endpoint: URLHelper::join() throws InvalidArgumentException when the endpoint is a valid absolute URL, unless explicitly allowed, requiring callers to opt-in to the functionality on a per-connector or per-request basis.