The Agent Skills Directory

[COMMAND_EXECUTION]: The reference file references/linux_commands.md documents high-risk network and command-line patterns, including reverse shell configurations such as nc -lvnp 4444 -e /bin/bash and socat TCP-LISTEN:4444,reuseaddr,fork EXEC:/bin/bash. These are presented as educational references for interacting with target systems in CTF environments.
[PRIVILEGE_ESCALATION]: The skill includes documentation for privilege escalation reconnaissance in references/linux_commands.md. It lists commands for identifying misconfigured permissions, such as sudo -l and searching for SUID binaries using find / -perm -4000. This information is consistent with the skill's stated purpose for security challenges.
[INDIRECT_PROMPT_INJECTION]: Several scripts and snippets in the skill (e.g., qr_scan_all.sh, spectrogram.sh, and bf_decode.py) are designed to ingest and process external, potentially untrusted data such as image files, audio files, and source code. This creates an attack surface for indirect prompt injection if an attacker crafts malicious file content intended to manipulate the agent's interpretation or output during analysis.

misc-tools