rsa-attack-techniques
Pass
Audited by Gen Agent Trust Hub on Jul 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides Python and SageMath implementations for well-known cryptographic attacks (e.g., Wiener's, Coppersmith's, Pollard's p-1). All code snippets are standard mathematical procedures for integer factorization and modular arithmetic.
- [SAFE]: No evidence of prompt injection or attempts to bypass AI safety guidelines. The 'AI LOAD INSTRUCTION' is used appropriately to define the agent's persona for the task.
- [SAFE]: There are no hidden network operations, data exfiltration patterns, or unauthorized file access. The referenced external services (such as factordb.com) and tools (RsaCtfTool, yafu, msieve) are standard in the security community.
- [SAFE]: No obfuscation techniques (e.g., Base64, zero-width characters, or homoglyphs) were detected in the instructions or scripts.
- [SAFE]: The skill does not perform any automated remote code execution or package installation. It provides command-line examples for the user to execute manually in their own environment.
Audit Metadata