The overly permissive sandbox configuration in DSPy allows attackers to steal sensitive files in cases when users build an AI agent which consumes user input and uses the “PythonInterpreter” class.
Advisories
Source ID Title
Github GHSA Github GHSA GHSA-vvw2-h478-xwr3 DSPy does not properly restrict file reads
Fixes

Solution

No solution given by the vendor.


Workaround

No workaround given by the vendor.

History

Wed, 05 Nov 2025 11:00:00 +0000

Type Values Removed Values Added
First Time appeared Jfrog
Jfrog dspy
Vendors & Products Jfrog
Jfrog dspy

Tue, 04 Nov 2025 15:15:00 +0000

Type Values Removed Values Added
Metrics ssvc

{'options': {'Automatable': 'yes', 'Exploitation': 'poc', 'Technical Impact': 'partial'}, 'version': '2.0.3'}


Tue, 04 Nov 2025 13:30:00 +0000

Type Values Removed Values Added
Description The overly permissive sandbox configuration in DSPy allows attackers to steal sensitive files in cases when users build an AI agent which consumes user input and uses the “PythonInterpreter” class.
Title Insecure configuration in DSPy lead to arbitrary file read when running untrusted code inside the sandbox
Weaknesses CWE-653
References
Metrics cvssV3_1

{'score': 5.9, 'vector': 'CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N'}


cve-icon MITRE

Status: PUBLISHED

Assigner: JFROG

Published:

Updated: 2025-11-04T14:19:33.945Z

Reserved: 2025-11-04T13:05:46.639Z

Link: CVE-2025-12695

cve-icon Vulnrichment

Updated: 2025-11-04T14:19:26.090Z

cve-icon NVD

Status : Awaiting Analysis

Published: 2025-11-04T14:15:34.087

Modified: 2025-11-04T15:40:45.533

Link: CVE-2025-12695

cve-icon Redhat

No data.

cve-icon OpenCVE Enrichment

Updated: 2025-11-05T10:47:25Z