content-evaluation-framework
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a content quality auditor that utilizes local reference files and structured rubrics to evaluate educational materials. No malicious behaviors or security vulnerabilities were identified within the analyzed scripts.
- [COMMAND_EXECUTION]: The skill makes use of standard file-utility tools (Read, Grep, Glob) for the purpose of inspecting content files and identifying pattern violations. These operations are restricted to the local file system and align with the skill's primary auditing function.
- [PROMPT_INJECTION]: The skill provides explicit instructions that establish rigid guardrails (e.g., 'NEVER pass content that violates constitutional principles', 'NEVER provide scores without evidence'). These instructions prioritize project safety guidelines over potentially conflicting user data.
- [DATA_EXFILTRATION]: No network capabilities or remote data transfer patterns were detected. Interactions are limited to reading local documentation and user-provided lesson drafts within the workspace environment.
Audit Metadata