Flowise Mass Assignment Vulnerability in Leads Endpoint Allowing Unauthenticated Database Injection

A mass assignment vulnerability has been identified in Flowise, a user interface for building customized large language model flows. This vulnerability, present in versions through 3.0.12, allows unauthenticated users to inject arbitrary values into internal database fields when creating leads. The issue arises because the application whitelists the '/api/v1/leads' endpoint, enabling unauthorized access. Exploitation involves manipulating fields that should be auto-generated by the server, such as 'id' and 'createdDate', as well as 'chatId'. The vulnerability has been patched in version 3.0.13.

Impact

Exploitation of this vulnerability could lead to unauthorized control over internal database fields, allowing for injection of arbitrary values. This could disrupt data integrity, manipulate audit trails, and cause conflicts within the database by overwriting existing records. Additionally, the vulnerability could be exploited to link leads to arbitrary chatflows, bypassing normal ownership rules.

Reproduction

The vulnerability can be reproduced by sending a POST request to the '/api/v1/leads' endpoint without authentication. The request body can include values for the 'id', 'createdDate', and 'chatId' fields, which will be accepted and processed by the application, bypassing the intended auto-generation of these fields.

Remediation

Users can update to Flowise version 3.0.13, where this vulnerability has been patched.