re-arm
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing standard reverse engineering tools, including binwalk and unblob via the Python package manager (pip), and ghidra, qemu, and gdb via system package managers like apt, brew, and pacman.\n- [COMMAND_EXECUTION]: The skill uses various command-line tools for binary analysis and environment verification, such as file, readelf, nm, and objdump.\n- [DYNAMIC_EXECUTION]: The skill includes instructions for dynamic analysis using QEMU (qemu-arm, qemu-aarch64) to emulate ARM binaries and GDB for debugging. It explicitly advises the use of sandboxing and network isolation for these activities to maintain security.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external binary files (target.bin, target.elf), creating a potential surface for indirect prompt injection if those files contain embedded instructions.\n
- Ingestion points: Input binary files (target.bin, target.elf) mentioned in SKILL.md.\n
- Boundary markers: No explicit boundary markers or warnings for untrusted data content are provided in the instructions.\n
- Capability inventory: The skill utilizes qemu for emulation, gdb for debugging, and ghidra or IDA for static analysis (SKILL.md).\n
- Sanitization: No sanitization of the input binary content is described, though the skill recommends execution within a sandbox.
Audit Metadata