state-invariant-detection
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill package is composed of documentation and instructional content intended for security auditing. Evidence: All files (SKILL.md, references/case-studies.md, and references/invariant-types.md) contain only text-based instructions and Solidity code examples.
- [NO_CODE]: No executable scripts or configuration files were found in the provided content. Evidence: There are no .py, .js, or binary files, nor are there package managers like package.json or requirements.txt.
- [PROMPT_INJECTION]: The instructions do not contain any patterns typical of prompt injection, such as commands to ignore previous rules, bypass safety filters, or extract system instructions.
- [DATA_EXFILTRATION]: No network operations, hardcoded credentials, or sensitive file path access patterns were identified in the analysis.
Audit Metadata