Security Flaw in PraisonAI Affects Recipe Execution Process
CVE-2026-57142

7.8HIGH

Key Information:

Vendor

Mervinpraison

Status
Praisonai
Vendor
CVE Published:
23 October 2026

What is CVE-2026-57142?

PraisonAI, a multi-agent system, has a critical vulnerability in its recipe execution policy that allows for untrusted commands to be executed. Specifically, the dangerous-tool policy intended to restrict access to default-denied tools can be bypassed. This occurs because the validation process only examines tools declared in TEMPLATE.yaml requires.tools, while allowing untrusted workflows to introduce shell commands via workflow.yaml agents.*.tools. As a result, recipes can be executed without proper validation, leading to potential risks in system integrity. This vulnerability was addressed in version 4.6.61.

Affected Version(s)

PraisonAI >= 4.5.87, < 4.6.61

References

https://github.com/MervinPraison/PraisonAI/security/advis...
x_refsource_CONFIRM

CVSS V3.1

Score:
7.8
Severity:
HIGH
Confidentiality:
High
Integrity:
High
Availability:
High
Attack Vector:
Local
Attack Complexity:
Low
Privileges Required:
None
User Interaction:
Required
Scope:
Unchanged

Timeline

  • Vulnerability published

  • Vulnerability Reserved

.