The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute various system inspection commands, including find, grep, lsof, pgrep, and launchctl, to audit the machine's state and configuration.\n- [DATA_EXFILTRATION]: The skill accesses sensitive local data by scanning the ~/Documents directory and reading shell configuration files such as ~/.zshrc and ~/.config. This constitutes significant data exposure; however, the instructions include a strict prohibition on network calls, which mitigates the risk of external exfiltration.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted data from user-provided security advisories and local project manifest files (package.json, requirements.txt, etc.). The agent uses these values to construct shell commands without explicit sanitization or boundary markers, creating a surface where malicious file content could influence command execution.

cyber-audit