implicit-assumption
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is purely instructional, providing a methodology for logical analysis based on the book 'First Principles' by Li Shanyou. It does not attempt to access the network, file system, or external resources.
- [NO_CODE]: No executable scripts, binaries, or shell commands are present in the skill files. All logic is implemented via natural language instructions and JSON-based test cases for agent triggering.
- [METADATA_POISONING]: The skill includes an 'Audit Information' section with a future date (2026). While this is a non-standard metadata field, it is used for internal documentation purposes and does not contain instructions to override safety filters or mislead the analyzer.
Audit Metadata