CVE-2026-57125 PraisonAI - Command Injection

PraisonAI < 4.6.59 and praisonaiagents < 1.6.59 contain a command injection caused by unauthenticated POST /api/v1/runs accepting attacker-controlled agent_yaml and bypassing approval checks, letting remote attackers execute arbitrary OS commands without credentials or interaction.

CVSS base score 9.8 /10
Critical
CISA KEV
Not listed
Public exploit
None known
Authentication
Not required
EPSS
0.60% chance of exploitation in 30 days
View the NVD record ↗

Details

Type
Command Injection
Impact
Remote Code Execution
CVSS vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Published
14 September 2026
Updated
15 October 2026
Weakness
CWE-306

Vendor fix

Update to praisonai 4.6.59 and praisonaiagents 1.6.59 or later.

Patching closes this CVE. It does not tell you which of your systems run the affected version, or what else is exposed alongside them.

References

← All CVEs