CVE-2026-56075: PraisonAI - Arbitrary Shell Command Execution via Hardcoded Approval Mode Override
PraisonAI before 4.5.128 contains an arbitrary shell command execution vulnerability where the UI modules hardcode approvalmode to auto, overriding administrator configuration from PRAISONAPPROVALMODE 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.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
PraisonAIto a version that resolves this vulnerability.Fixed in 4.5.128 - Configuration
Update PraisonAI so the UI modules no longer hardcode approval_mode to auto in versions before 4.5.128; ensure approval_mode follows the administrator configuration specified by the PRAISON_APPROVAL_MODE environment variable.
PraisonAI UI modules approval_mode = auto (hardcoded in vulnerable versions; override to administrator-configured value via PRAISON_APPROVAL_MODE)
Event History
Frequently Asked Questions
What is the severity of CVE-2026-56075?
The severity of CVE-2026-56075 is high, with a CVSS score of 8.8.
What kind of vulnerability is CVE-2026-56075?
CVE-2026-56075 is an arbitrary shell command execution vulnerability found in PraisonAI.
How do I fix CVE-2026-56075?
To fix CVE-2026-56075, upgrade to PraisonAI version 4.5.128 or later.
Who is affected by CVE-2026-56075?
Authenticated attackers can exploit CVE-2026-56075 if they have access to the PraisonAI application.
What does CVE-2026-56075 exploit in PraisonAI?
CVE-2026-56075 exploits a hardcoded approval mode that overrides administrator configuration, enabling arbitrary command execution.