Impact
Guardrails AI versions up to and including 0.6.7 allow an attacker who can publish a package to the Hub to inject arbitrary code during the installation of that package. The vulnerability arises in the Hub package installation mechanism where manifest data is used to construct a script path that is then executed without validation or sanitization. This flaw is classified as CWE-94, enabling remote execution of malicious code by any system that installs the compromised package.
Affected Systems
The vulnerability affects Guardrails AI’s Hub installation process for package versions 0.6.7 and earlier, with no specific version sublist provided. Any environment that uses guardrails AI to install packages from the Hub is potentially impacted.
Risk and Exploitability
The CVSS score is not supplied, but the impact is severe because it permits execution of arbitrary code on the target system. EPSS data is unavailable, and the vulnerability is not listed in the CISA KEV catalog. The likely attack vector is an attacker who can publish or push a malicious package to the Hub; once a user installs that package, the crafted post_install script runs with no restrictions. The overall risk is high, with the potential for complete compromise of the affected machine.
OpenCVE Enrichment