The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a wide array of built-in Linux utilities to audit system state, including find, grep, awk, systemctl, rpm, dpkg, and bpftool. These commands are used to enumerate services, timers, kernel modules, and package integrity.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze data from a potentially compromised host, such as the contents of service files, crontabs, and shell initialization scripts. This creates a surface where an attacker who has already compromised the target system could place malicious instructions in these files to attempt to influence the agent's analysis.
Ingestion points: Reads contents from files in /etc/systemd/, /etc/cron*, ~/.bashrc, ~/.ssh/authorized_keys, and other configuration directories.
Boundary markers: None identified. The skill does not instruct the agent to use specific delimiters or ignore embedded instructions when reading file content.
Capability inventory: The skill possesses extensive read access to system files and diagnostic tools, but lacks network exfiltration or file-writing capabilities.
Sanitization: No specific sanitization or filtering of the ingested file content is performed before the agent processes it.

analyzing-linux-persistence