seija-assumption-inversion
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and does not contain any executable code, network operations, or sensitive data access.
- [SAFE]: The references to external websites (official Touhou Project pages by ZUN) are for provenance and historical context only and are well-known in the relevant community.
- [SAFE]: The 'evals/cases.json' file includes safety test cases designed to ensure the agent does not use the skill to deny documented security incidents or manufacture controversy.
Audit Metadata