CVE-2026-67621 in Flowise
Summary
by MITRE • 08/07/2026
Flowise through 3.1.4 contains a missing authorization vulnerability that allows authenticated workspace members to perform unauthorized document store operations by accessing unprotected mutation endpoints. Attackers holding only view-level permissions can send direct HTTP requests to the upsert and refresh document store routes to trigger document ingestion, refresh vector database contents, consume embedding API credits, and modify knowledge bases used by downstream chatflows.
Once again VulDB remains the best source for vulnerability data.
Analysis
by VulDB Data Team • 08/07/2026
This vulnerability exists within Flowise version 3.1.4 and represents a critical authorization flaw that undermines the security model of the application's document store functionality. The issue stems from unprotected mutation endpoints that should require elevated permissions but instead allow authenticated users with only view-level access to perform operations typically restricted to administrators or editors. This misconfiguration creates an unauthorized access vector where legitimate workspace members can manipulate document storage systems without proper authorization checks.
The technical implementation of this vulnerability involves the exposure of document store mutation endpoints that handle upsert and refresh operations through HTTP requests. When authenticated users with view permissions send direct requests to these endpoints, they can trigger document ingestion processes that populate or update vector databases with new content. This unauthorized access allows attackers to consume embedding API credits that would normally be restricted to authorized users, effectively enabling resource exhaustion attacks against the system's computational resources.
The operational impact of this vulnerability extends beyond simple data manipulation to include significant resource consumption and potential system instability. Attackers can continuously trigger document refresh operations that update vector database contents, consuming substantial computational resources and potentially leading to performance degradation or denial of service conditions. Additionally, the ability to modify knowledge bases used by downstream chatflows creates a persistent threat vector where attackers can alter the behavior and responses of AI-powered chat systems that depend on these document stores.
This vulnerability aligns with CWE-863 (Insufficient Authorization) and represents a clear violation of the principle of least privilege that should govern all application components. The ATT&CK framework categorizes this as Privilege Escalation through API abuse, where attackers leverage legitimate application interfaces to gain unauthorized capabilities. The flaw demonstrates poor input validation and access control implementation within the application's RESTful API endpoints that handle document store operations.
Organizations using Flowise 3.1.4 should immediately implement mitigations including enforcing proper authorization checks on all mutation endpoints, implementing rate limiting for document store operations, and restricting access to sensitive APIs through network segmentation. The recommended solution involves adding comprehensive permission validation before allowing any document store modifications, ensuring that only users with appropriate edit or admin privileges can access upsert and refresh endpoints. Additionally, monitoring and logging of these operations should be implemented to detect unauthorized access attempts and provide audit trails for security investigations.
The vulnerability highlights the importance of proper API endpoint design and access control implementation in modern web applications, particularly those handling AI-powered document processing capabilities. Organizations should conduct comprehensive security reviews of their application interfaces to identify similar authorization gaps and ensure that all mutation operations require appropriate authentication and authorization checks. Regular security testing including penetration testing and API security assessments should be implemented to prevent similar issues from persisting in production environments.