Mitigation Strategies

To mitigate this vulnerability, upgrade PraisonAIAgents to version 1.5.128 or later, where the issue with read_skill_file() has been fixed.

Useful 0 Not Useful 0

Impact Analysis

This vulnerability can lead to unauthorized disclosure of sensitive files from the filesystem. Since an attacker can exploit the read_skill_file() function to read arbitrary files without any approval, sensitive information could be exfiltrated, potentially leading to data breaches or exposure of confidential data.

Useful 0 Not Useful 0

Executive Summary

The vulnerability exists in PraisonAIAgents, a multi-agent teams system, specifically in versions prior to 1.5.128. The function read_skill_file() in skill_tools.py allows reading arbitrary files from the filesystem because it accepts an unrestricted skill_path parameter. Unlike other functions that enforce workspace boundaries or require approval, read_skill_file() has no such protections. This means an agent influenced by prompt injection can read and exfiltrate sensitive files without any approval prompt.

Useful 0 Not Useful 0

Compliance Impact

This vulnerability allows an agent to read arbitrary files from the filesystem without any approval prompt, potentially leading to unauthorized exfiltration of sensitive data.

Such unauthorized access and exfiltration of sensitive information could lead to non-compliance with data protection regulations like GDPR and HIPAA, which require strict controls on access to and disclosure of personal and sensitive data.

Useful 0 Not Useful 0