gaokao-politics-question-generator

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [Indirect Prompt Injection] (SAFE): The skill defines a surface for processing external exam questions against pedagogical standards. This is a low-risk data ingestion point. Evidence: 1. Ingestion points: The skill uses reference documents in the /references/ directory to analyze or generate questions. 2. Boundary markers: The module files include explicit 'Out-of-scope warnings' (超纲预警) to prevent the AI from generating inappropriate or excessive content. 3. Capability inventory: No command execution or network capabilities were found in the skill files. 4. Sanitization: Not applicable for these static reference files.\n- [NO_CODE] (SAFE): The skill contains no shell scripts, Python files, JavaScript, or binaries. All files are purely informational Markdown. No package manifests (e.g., package.json, requirements.txt) are present.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:35 PM