Skip to content

Security Findings in test_gemma4_local.py #330

Description

@joshua-trustabl

We ran the Trustabl scanner on your repo and identified some security findings. In scripts/test_gemma4_local.py, we found that LlmAgent has no description, which means Google ADK routes delegation between agents using the description= field on each LlmAgent. Additionally, Agent has no safety_settings, indicating that Google ADK agents use Gemini models whose configurable content filters default to OFF when safety_settings is not provided. Both findings are flagged as MEDIUM severity. We wanted to point these out in case they were unintentional omissions.


Add Trustabl to your CI — trustabl/trustabl-action:

- name: Trustabl scan
  uses: trustabl/trustabl-action@v1

https://trustabl.ai

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions