ageval-config-package
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFE
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill provides explicit instructions to avoid hardcoding secrets in YAML files or container layers, recommending environment variable locators for sensitive data like API keys.
- [INDIRECT_PROMPT_INJECTION]: Documentation mandates that 'gold' evaluation data be stored in paths isolated from the agent's workspace, creating a security boundary that prevents unauthorized access during execution.
- [REMOTE_CODE_EXECUTION]: The framework encourages baking dependencies into immutable container images rather than using runtime package installations, which mitigates risks associated with dynamic dependency fetching.
Audit Metadata