lattice-crypto-attacks
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides educational content and code snippets for cryptographic analysis in CTF and research contexts.
- [DATA_EXFILTRATION]: Analysis of the skill body shows no network operations (curl, wget) or access to sensitive system file paths (e.g., .ssh, .aws, .env).
- [PROMPT_INJECTION]: The instructions do not contain any patterns intended to bypass safety filters, extract system prompts, or override agent constraints.
- [REMOTE_CODE_EXECUTION]: The skill includes Python and SageMath code snippets for mathematical computations, but does not download or execute code from remote or untrusted sources.
- [OBFUSCATION]: No hidden content, encoded strings, or homoglyph attacks were detected in the text or metadata.
Audit Metadata