CVE-2026-56075
HighCVSS 8.8Exploitation Probability (EPSS)
Low risk37th percentile - higher than 37% of all known CVEs
Summary
PraisonAI before 4.5.128 contains an arbitrary shell command execution vulnerability where the UI modules hardcode approval_mode to auto, overriding administrator configuration from PRAISON_APPROVAL_MODE environment variable. Authenticated attackers can instruct the LLM agent to execute arbitrary shell commands.
Risk Assessment
This vulnerability allows attackers to bypass approval mechanisms, potentially leading to unauthorized access and damage to the system. Organizations may face serious data security threats.
Recommendation
It is recommended to update PraisonAI to version 4.5.128 or later to eliminate this vulnerability. Additionally, conducting an audit of environmental configurations to secure environment variables is advisable.
Original NVD description (English source)
PraisonAI before 4.5.128 contains an arbitrary shell command execution vulnerability where the UI modules hardcode approval_mode to auto, overriding administrator configuration from PRAISON_APPROVAL_MODE environment variable. Authenticated attackers can instruct the LLM agent to execute arbitrary shell commands via subprocess.run with shell=True, bypassing the manual approval gate and insufficient command sanitization blocklists.