framework-selection
Pass
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [Architectural Guidance]: The skill provides high-level logic and decision tables to help an agent choose between different framework layers. This is purely informational and does not perform any technical operations.
- [No Executable Content]: An analysis of the skill reveals no scripts, shell commands, or package installations. Since the skill consists entirely of markdown-based instructions, it does not pose a direct execution risk.
- [Capability Awareness]: The text mentions framework features like 'FilesystemMiddleware' and 'SkillsMiddleware'. These are described as part of the Deep Agents framework's built-in capabilities. While these features facilitate file access or skill loading in a project, this skill itself does not implement them or process untrusted data through them.
Audit Metadata