CVE-2026-101861 describes an eval injection vulnerability in langflow-ai's langflow product affecting versions >=1.0.16 <1.12.0 and >=0.0.94 <1.12.0. The flaw exists in schema.py where an eval() call is used unsafely during the conversion of components into LangChain tools via ComponentToolkit.get_tools(). Authenticated attackers can place Python objects with malicious __repr__ methods into component input options lists, triggering arbitrary code execution when eval() processes these options interpolated into a Literal type string. This unsafe eval usage lacks proper sanitization or safe evaluation mechanisms.

The vulnerability allows authenticated attackers with high privileges to execute arbitrary code on the system running langflow by exploiting the unsafe eval() call. The CVSS score is low (2.1) due to the requirement for high privileges and user interaction, and limited scope and impact. There are no known exploits in the wild. The impact is limited to authenticated users who can manipulate component input options.

No official patch or remediation information is provided in the input data. Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until a fix is available, restrict access to authenticated users and avoid using vulnerable versions if possible.