CrewAI Framework contains a flaw in the load_agent_from_repository function that fails to restrict a user‑supplied argument before using it to import a module. This unsafe reflection vulnerability allows an attacker to execute arbitrary code in the context of the service account by providing a malicious agent configuration via the repository. The flaw falls under CWE‑470, a classic example of dynamic code execution based on untrusted input.

The affected vendor is CrewAI, and the product impacted is the CrewAI Framework. No specific version numbers were supplied in the advisory, so all installations of the framework may be vulnerable until a fix is applied.

The CVSS score of 8.8 classifies the issue as a high‑severity remote code execution risk. The EPSS score is not available, but the advisory notes that user interaction is required: the target must load a malicious agent configuration from the repository. The vulnerability is not listed in the CISA KEV catalog. Because the flaw permits code execution with the privileges of the service account, successful exploitation would grant an attacker full control over the affected system.