PraisonAIAgents, a multi-agent system, had a vulnerability in versions before 1.5.128 where the execute_command function in shell_tools.py used os.path.expandvars() on command arguments while running commands with shell=False. This manual environment variable expansion bypassed intended security controls, allowing secrets stored in environment variables to be exfiltrated. The approval system's display of unexpanded variables created a deceptive approval process, as the command shown to reviewers differed from the executed command. The vulnerability is classified under CWE-526 (Cleartext Storage of Sensitive Information) and carries a CVSS 3.1 score of 7.4 (High severity). The issue is resolved in version 1.5.128.

An attacker or user able to influence command arguments could cause sensitive environment variable values (e.g., database credentials, API keys) to be exposed or exfiltrated. The deceptive approval display increases risk by hiding the true command executed, potentially leading to unauthorized disclosure of secrets. There is no indication of impact on integrity or availability. No known exploits in the wild have been reported.

Upgrade PraisonAIAgents to version 1.5.128 or later, where this vulnerability is fixed. Since the vendor advisory indicates the issue is resolved in this version, applying this official fix is the recommended remediation. Patch status is not explicitly stated beyond the fix in 1.5.128, so verify with the vendor for any additional guidance.