CVE-2026-42079 PPTAgent: Arbitrary Code Execution via Python eval() of LLM-Generated Code with Builtins in Scope
Attackers can execute arbitrary code by exploiting unsafe eval() usage on LLM-generated content in the PPTAgent library.
- Exploited in the wild
- Not known
- Login needed
- Not needed
- Attackable
- Local or same network only
- Victim action needed
- A user has to click or open something
- Fix
- Available
What is CVE-2026-42079?
CVE-2026-42079 is a high-severity code injection vulnerability in icip-cas PPTAgent, published on 4 May 2026. It can only be exploited from the same machine or local network, and only if a user opens a file or follows a link.
Official description
PPTAgent is an agentic framework for reflective PowerPoint generation. Prior to commit 418491a, PPTAgent is vulnerable to arbitrary code execution via Python eval() of LLM-generated code with builtins in scope. This issue has been patched via commit 418491a.
Weakness: CWE-95
Which versions are affected by CVE-2026-42079?
3 open source packages ship an affected version, including pptagent.
| Package | Affected | Fixed in |
|---|---|---|
pptagent PyPI | All versions | 1.1.36 |
pptagent PyPI | All versions | 1.1.36 |
pptagent PyPI | All versions | 1.1.36 |
How do I fix CVE-2026-42079?
Upgrade pptagent to 1.1.36 or later.
Upgrade, and never evaluate templates, expressions or code built from user input. More about Code injection.
Is CVE-2026-42079 being exploited?
It is not confirmed as exploited in the wild, but public exploit code for CVE-2026-42079 exists.
Sources
Also known as GHSA-89g2-xw5c-v95p, PYSEC-2026-2892.