agentos-api-evals
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill is a legitimate integration for the AgentOS platform. It follows standard patterns for API interaction and includes placeholders for credentials rather than hardcoded secrets.
- [EXTERNAL_DOWNLOADS]: The skill specifies a dependency on the
agnolibrary (formerly Phidata), which is a well-known and reputable framework for agent orchestration and evaluation. The dependency is installed via standard package management (uv/pip). - [COMMAND_EXECUTION]: The skill provides a utility script,
scripts/run_evals.py, designed to facilitate evaluation tasks. The script performs transparent network operations to a user-defined or local endpoint and does not contain any hidden or unauthorized system commands.
Audit Metadata