hosted-agents
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides educational content and implementation patterns for building secure agent infrastructure. All external references are to well-known, trusted services including Cloudflare, Modal, GitHub, and official developer documentation.
- [SAFE]: Code examples in
references/infrastructure-patterns.mdandscripts/sandbox_manager.pydemonstrate sandbox isolation and least-privilege principles, such as using short-lived GitHub App installation tokens and restricting network access within sandboxes. - [SAFE]: The skill correctly handles user identity for git attribution, ensuring that commits are signed by the prompting user rather than the application service account.
- [SAFE]: Implementation patterns for Cloudflare Durable Objects and Modal Sandboxes follow industry best practices for session isolation and state management.
Audit Metadata