CVE-2026-34937

What it is

PraisonAI is a multi-agent teams system. Prior to version 1.5.90, run_python() in praisonai constructs a shell command string by interpolating user-controlled code into python3 -c "<code>" and passing it to subprocess.run(..., shell=True). The escaping logic only handles \ and ", leaving $() and backtick substitutions unescaped, allowing arbitrary OS command execution before Python is invoked. This issue has been patched in version 1.5.90.

At a glance

Exploitability

Can be exploited locally on the machine with a low-privilege account. High risk of data exposure, data tampering, service disruption.

Exploit odds

0.6% chance of exploitation in the next 30 days (EPSS).

Fix

A patch, advisory or fixed version is referenced. Upgrade or apply it.

Timeline

1. 2026-04-03 23:17 UTC PublishedDisclosed publicly on CVE.org and NVD
2. 2026-07-24 22:10 UTC Last updatedMost recent change to the record

Why it ranks here

• Exploited in the wild: no, CISA does not list it.
• Exploit odds: 0.6% in the next 30 days (EPSS).
• Severity: 7.8 HIGH (CVSS).
• Fix: a patch or advisory is linked.

Details

Published

2026-04-03 23:17 UTC

Last modified

2026-07-24 22:10 UTC

Exploit probability

0.6% · EPSS, next 30 days

Weakness

CWE-78 · OS command injection

Record status

Analyzed

Affected

• Praison Praisonaiagents < 1.5.90

Attack profile CVSS 3.1 · 7.8

Attack vector

Local

Complexity

Low

Privileges needed

Low

User interaction

None

Scope

Unchanged

Confidentiality

High

Integrity

High

Availability

High

Is it in your code?

Hand this to your coding agent. It searches your project for the affected products and checks versions. Nothing leaves your machine.

Read https://cve-list.com/agent/cve/CVE-2026-34937.md and follow it. Check whether this project is affected by CVE-2026-34937. Read only. Don't change anything and don't upload my code. Tell me yes, no or not sure, with the file as proof.

Copy as ticket · More options · Status badge