PraisonAI, a multi-agent teams system by MervinPraison, contains a code injection vulnerability (CWE-94) in versions before 4.5.115. The create_agent_centric_tools() function returns tools such as acp_create_file that process file content through template rendering. When user input provided via agent.start() is passed directly to these tools without escaping, template expressions embedded in the input are executed, enabling code injection. This vulnerability has a CVSS 3.1 score of 8.8, indicating high impact on confidentiality, integrity, and availability. The vulnerability is resolved in version 4.5.115. There is no vendor advisory or patch link provided, and no known exploits have been reported.

Successful exploitation allows an attacker with low privileges and no user interaction to execute arbitrary code via template injection, potentially leading to full compromise of the affected system's confidentiality, integrity, and availability. This can result in unauthorized code execution and system control.

Upgrade PraisonAI to version 4.5.115 or later, where this vulnerability is fixed. Since no official patch link or vendor advisory is provided, users should verify the version and apply the update accordingly. Patch status is not yet confirmed by a vendor advisory; check for official updates from MervinPraison. Until patched, avoid passing unescaped user input to the create_agent_centric_tools() function.