CVE-2026-78379: CWE-1427: Improper Neutralization of Input Used for LLM Prompting in Amazon strands-agents-tools

The vulnerability CVE-2026-78379 affects the python_repl tool in Amazon Strands Agents Tools before version 0.8.5. It is caused by improper neutralization of input used for LLM prompting, specifically when a crafted prompt forwards the non_interactive_mode keyword argument through the batch tool. This enables remote actors to bypass the human consent gate and execute arbitrary Python code on the host system. The CVSS 4.0 score is 9.2, indicating a critical severity with network attack vector, low attack complexity, no privileges required, no user interaction, and high impact on confidentiality, integrity, and availability. The vendor advisory from AWS is available but does not explicitly state the remediation level or patch availability in the provided data.

Successful exploitation allows remote attackers to execute arbitrary Python code on the host running the vulnerable agent without requiring user interaction or privileges. This can lead to full compromise of the affected system, including unauthorized access, data manipulation, or disruption of service.

Users should upgrade Amazon Strands Agents Tools to version 0.8.5 or later to remediate this vulnerability. The vendor advisory should be consulted for the latest remediation guidance. Patch status is not explicitly confirmed in the provided data, but the description clearly states upgrading to 0.8.5 or later as the fix.