Langflow /api/v1/validate/code Arbitrary Code Execution Vulnerability and Fix

Vulnerability Overview A code injection vulnerability exists in the code validation endpoint of Langflow. This endpoint accepts arbitrary Python code from authenticated users and executes it server-side using the function without adequate sandboxing or security controls. This allows authenticated attackers to execute arbitrary Python code on the server, resulting in complete system compromise. Impact Scope Read/Write arbitrary files on the server Execute system commands Access environment variables and secrets Establish reverse shells Lateral movement within the network Remediation 1. Do not directly execute user-provided code: Code validation should be performed solely through static analysis (e.g., AST parsing, linting). 2. Implement proper sandboxing: If dynamic execution is absolutely necessary, use a restricted execution environment, such as: - Docker containers with limited capabilities - Python’s library - Isolated sandbox processes with resource limits 3. Apply the principle of least privilege: Validation should not require full code execution capabilities. 4. Add authorization checks: Restrict this functionality to administrator users only (if the feature must exist). 5. Input validation: Reject code containing dangerous imports or operations. POC Code Remediation Code Fixed Version Langflow Version: 1.9.3 Python Version: 3.10 Operating System: Ubuntu 22.04 Fix Status Fix Commit: rsd-darshan submitted the fix on June 1, 2024. Fix Merge: The fix has been merged into the branch, and the issue has been closed. Fix Details Fix Content: The function was modified to compile nodes only (to capture syntax/compilation errors) instead of executing , thereby preventing the execution of user-provided code. Regression Testing: Regression tests were added covering default arguments, module-level expressions, class attributes, and decorator expressions. Post-Fix Behavior Code Validation Endpoint: No longer executes user-provided code; performs only static analysis. Security Controls: Implemented proper sandboxing and the principle of least privilege to ensure code validation security. Post-Fix Verification Verification Steps: 1. Install Langflow version 1.9.3 or higher. 2. Start the Langflow server. 3. Create a new Flow and add custom Python code. 4. Attempt to inject malicious code. 5. Verify that the code is correctly rejected or subjected only to static analysis. Post-Fix Expected Behavior Code Validation Endpoint: No longer executes user-provided code; performs only static analysis. Security Controls: Implemented proper sandboxing and the principle of least privilege to ensure code validation security. Post-Fix Verification Verification Steps: 1. Install Langflow version 1.9.3 or higher. 2. Start the Langflow server. 3. Create a new Flow and add custom Python code. 4. Attempt to inject malicious code. 5. Verify that the code is correctly rejected or subjected only to static analysis. Post-Fix Expected Behavior Code Validation Endpoint: No longer executes user-provided code; performs only static analysis. Security Controls: Implemented proper sandboxing and the principle of least privilege to ensure code validation security.

Goal Reached Thanks to every supporter — we hit 100%!

Langflow /api/v1/validate/code Arbitrary Code Execution Vulnerability and Fix

More from this source