code-obfuscation-deobfuscation
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides procedures for analyzing and reverse-engineering untrusted binary data. This represents an attack surface where instructions hidden within analyzed binaries (e.g., in strings, bytecode, or metadata) could attempt to influence the agent's behavior. The skill mitigates this by providing structured analysis templates and recommending specialized tools for emulation and symbolic execution.
- [COMMAND_EXECUTION]: The skill recommends the use of professional reverse engineering and analysis tools, including IDA Pro, Ghidra, x64dbg, and binary analysis frameworks. These recommendations are consistent with the skill's primary purpose of code deobfuscation analysis.
- [DYNAMIC_EXECUTION]: The provided Python snippets demonstrate the use of standard libraries like Z3 for theorem proving and Unicorn for CPU emulation. These are standard techniques in the security research community for analyzing complex code behaviors in a controlled environment.
Audit Metadata