Home

Description

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.

PUBLISHED Reserved 2026-06-18 | Published 2026-06-18 | Updated 2026-06-18 | Assigner VulnCheck




HIGH: 8.7CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
HIGH: 8.8CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

Problem types

Incorrect Authorization

Product status

Default status
unaffected

Any version before 4.5.128
affected

4.5.128 (semver)
unaffected

Credits

offset reporter

References

github.com/...isonAI/security/advisories/GHSA-qwgj-rrpj-75xm (GHSA Advisory GHSA-qwgj-rrpj-75xm) vendor-advisory

www.vulncheck.com/...on-via-hardcoded-approval-mode-override (VulnCheck Advisory: PraisonAI - Arbitrary Shell Command Execution via Hardcoded Approval Mode Override) third-party-advisory

cve.org (CVE-2026-56075)

nvd.nist.gov (CVE-2026-56075)

Download JSON