skyrim-modding-facts
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structured guidelines to prevent hallucination by prioritizing evidence from local files and official documentation. It explicitly restricts the agent from launching GUI tools, which reduces the risk of unauthorized system interaction.
- [SAFE]: All external URLs point to well-known community modding projects on GitHub (such as powerof3, TES5Edit, and Mutagen) or to official documentation for AI agent platforms (including OpenAI, Anthropic, and Vercel). These are documented neutrally as legitimate references.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies ingestion points for untrusted data from the user's modding environment, including
ENVIRONMENT.md,TASK.md, and log files likeskse64.log. While this constitutes an attack surface, the instructions specify read-only access for factual verification, and the behavior is inherent to the skill's primary purpose of environment analysis.
Audit Metadata