CVE-2026-39888
PraisonAI is a multi-agent teams system. Prior to 1.5.115, execute_code() in praisonaiagents.tools.python_tools defaults to sandbox_mode="sandbox", which runs user code in a subprocess wrapped with a restricted __builtins__ dict and an AST-based blocklist. The AST blocklist embedded inside the subprocess wrapper (blocked_attrs of python_tools.py) contains only 11 attribute names — a strict subset of the 30+ names blocked in the direct-execution path. The four attributes that form a frame-traversal chain out of the sandbox are all absent from the subprocess list (__traceback__, tb_frame, f_back, and f_builtins). Chaining these attributes through a caught exception exposes the real Python builtins dict of the subprocess wrapper frame, from which exec can be retrieved and called under a non-blocked variable name — bypassing every remaining security layer. This vulnerability is fixed in 1.5.115.
Scoring
- Severity
- CRITICAL
- CVSS base score
- 10
- CVSS vector
- CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H
- EPSS probability
- 0.48%
- CWE
- CWE-657, CWE-693
- Published
- 2026-04-08
- Last modified
- 2026-04-10
Affected products
- MervinPraison praisonaiagents
Weakness type
Related vulnerabilities
- CVE-2026-48399 — Adobe Campaign Classic (ACC) | Violation of Secure Design Principles (CWE-657)
- CVE-2026-30792 — RustDesk Client Blindly Merges Unauthenticated Strategy Payloads, Bypassing Local Security Settings
- CVE-2025-54255 — Acrobat Reader | Violation of Secure Design Principles (CWE-657)
- CVE-2025-24887 — OpenCTI bypass of protected attribute update
- CVE-2024-57957 — Vulnerability of improper log information control in the UI framework module...
- CVE-2024-26139 — OpenCTI Authenticated Privilege Escalation
- CVE-2023-29320 — ZDI-CAN-20712: Adobe Acrobat Blacklist Bypass Design flaw
- CVE-2022-30683 — AEM Violation of Secure Design Principles Security feature bypass