principle-guard-the-context-window
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides documentation on how to manage the context window efficiently by isolating large payloads and being selective with data reads. It does not contain any executable scripts, command line operations, network requests, or sensitive file access. The configuration in agents/openai.yaml is standard and restricts implicit invocation, further limiting its risk profile.
Audit Metadata