A critical remote code execution vulnerability in Langflow is currently being actively exploited in the wild, posing an immediate risk to all unpatched instances.

This vulnerability occurs in the POST /api/v1/build_public_tmp/{flow_id}/flow endpoint, which allows unauthenticated attackers to supply arbitrary Python code within the data parameter. This input is processed by the exec() function without sandboxing, leading to full remote code execution on the underlying host.

The CVSS score of 9.5 reflects the critical severity of this flaw, as it allows complete system compromise without requiring any user credentials. Successful exploitation could lead to full unauthorized access to the host environment, data exfiltration, or the deployment of persistent malware, resulting in significant operational downtime and potential loss of sensitive intellectual property or customer data.

Immediate Action: Upgrade Langflow to version 1.9.0 or later immediately to incorporate the necessary input sanitization and logic changes.

Proactive Monitoring: Monitor server logs for suspicious POST requests to the /api/v1/build_public_tmp/ endpoint, specifically looking for payloads containing Python syntax or unusual data structures.

Compensating Controls: Deploy a Web Application Firewall (WAF) with rules designed to inspect and block malicious input containing Python keywords or code injection patterns in API requests.

Public Exploit Available: Yes, a weaponized exploit exists (a Metasploit module and ExploitDB entry are available).

Given the confirmed active exploitation and the availability of weaponized exploit code, this vulnerability represents an extreme risk to organizational security. Administrators must prioritize the deployment of version 1.9.0 immediately. If an immediate update is not feasible, the affected endpoint should be restricted or disabled to prevent unauthorized remote execution.