Vulnerability · Published Apr 3, 2026 · Assigned by GitHub_M
CVE-2026-34938PraisonAI: Python Sandbox Escape via str Subclass startswith() Override in execute_code
Critical 10.0
In plain English
A critical-severity security flaw in PraisonAI from MervinPraison. Proof-of-concept exploit code is public. The chance it is exploited in the next 30 days is estimated at 13.4% (higher than 96% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.
Official description
PraisonAI is a multi-agent teams system. Prior to version 1.5.90, execute_code() in praisonai-agents runs attacker-controlled Python inside a three-layer sandbox that can be fully bypassed by passing a str subclass with an overridden startswith() method to the _safe_getattr wrapper, achieving arbitrary OS command execution on the host. This issue has been patched in version 1.5.90.
Affected packages & fixed versions
| Ecosystem | Package | Fixed in | Advisory |
|---|---|---|---|
| PyPI | praisonaiagents | 1.5.90 | GHSA-6vh2-h83c-9294 |
| PyPI | praisonaiagents | 1.5.90 | PYSEC-2026-484 |
Affected products (as reported)
| Vendor | Product | Versions |
|---|---|---|
| MervinPraison | PraisonAI | < 1.5.90 |
Weakness: CWE-693 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H