supervisor-api-background-mode
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill utilizes official libraries (
databricks-sdk,databricks-openai, andmlflow) for all core functionality. - [SAFE]: Authentication instructions leverage standard Databricks CLI environment variables, ensuring secrets are handled through secure local configuration rather than hardcoded values.
- [SAFE]: The implementation in
agent_server/utils.pyuses standard asynchronous patterns for polling and status checks against the Databricks API. - [PROMPT_INJECTION]: As an AI agent framework, the skill inherently processes untrusted user input. However, it relies on the underlying Supervisor API's security model and does not introduce specific vulnerabilities or bypasses.
Audit Metadata