A critical vulnerability in PraisonAI allows unauthenticated remote attackers to execute arbitrary agent tasks and gain unauthorized access to service credentials.

The application fails to implement authentication or authorization checks on the API endpoint /api/v1/runs, permitting unauthenticated attackers to submit malicious prompts, manage jobs, and exfiltrate sensitive service credentials.

The absence of authentication on core API functions creates a direct path for attackers to hijack the multi-agent system, leading to full compromise of configured agent capabilities and connected service accounts. With a CVSS score of 9.8, this flaw presents a critical risk of data exfiltration and unauthorized system control, which could result in severe reputational damage and operational disruption.

Immediate Action: Update the PraisonAI package to version 4.6.59 or later immediately to enforce required authentication and authorization controls.

Proactive Monitoring: Review API access logs for anomalous requests to the /api/v1/runs endpoint, specifically looking for unexpected job submission patterns or unauthorized administrative actions.

Compensating Controls: Implement strict network access controls or a Web Application Firewall (WAF) to restrict access to the /api/v1/runs endpoint to trusted IP addresses only until the patch can be applied.

Public Exploit Available: Unknown

This vulnerability represents a critical security failure due to the total lack of authentication on sensitive API endpoints. Organizations currently running affected versions of PraisonAI must prioritize an immediate update to version 4.6.59. Failure to patch this issue leaves the environment susceptible to remote code execution and unauthorized data access by any actor with network visibility to the service.