CVE-2026-78379
Published:August 25, 2026
Updated:September 24, 2026
Improper neutralization of input used for LLM prompting in the python_repl tool in Amazon Strands Agents Tools before 0.8.5 might allow remote actors to execute arbitrary Python code on the agent's host by bypassing the human consent gate, via a crafted prompt that forwards non_interactive_mode as a keyword argument through the batch tool. To remediate this issue, users should upgrade to version 0.8.5 or later.
Affected Packages
strands-agents-tools (CONDA):
Affected version(s) >=0.1.5 <0.8.5Fix Suggestion:
Update to version 0.8.5https://github.com/strands-agents/tools.git (GITHUB):
Affected version(s) >=v0.1.0 <v0.8.4Fix Suggestion:
Update to version v0.8.4strands-agents-tools (PYTHON):
Affected version(s) >=0.0.1 <0.8.5Fix Suggestion:
Update to version 0.8.5Related Resources (2)
Do you need more information?
Contact UsCVSS v4
Base Score:
9.2
Attack Vector
NETWORK
Attack Complexity
LOW
Attack Requirements
PRESENT
Privileges Required
NONE
User Interaction
NONE
Vulnerable System Confidentiality
HIGH
Vulnerable System Integrity
HIGH
Vulnerable System Availability
HIGH
Subsequent System Confidentiality
NONE
Subsequent System Integrity
NONE
Subsequent System Availability
NONE
CVSS v3
Base Score:
8.1
Attack Vector
NETWORK
Attack Complexity
HIGH
Privileges Required
NONE
User Interaction
NONE
Scope
UNCHANGED
Confidentiality
HIGH
Integrity
HIGH
Availability
HIGH
Weakness Type (CWE)
Improper Neutralization of Input Used for LLM Prompting
EPSS
Base Score:
0.32