The Agent Skills Directory

[COMMAND_EXECUTION] (LOW): The skill contains scripts designed to audit system security which execute commands via subprocesses.
Evidence in scripts/security-checker.sh: Executes ufw, firewall-cmd, and auditctl to check system hardening status.
Evidence in scripts/vulnerability-scanner.py: Executes pip-audit and npm audit to scan dependencies for known vulnerabilities.
[EXTERNAL_DOWNLOADS] (LOW): Utility scripts perform outbound network requests to validate security configurations of remote servers.
Evidence in scripts/headers-validator.sh: Uses curl to fetch and inspect HTTP headers.
Evidence in scripts/vulnerability-scanner.py: Uses urllib.request to check for security headers and TLS configuration.
[DATA_EXPOSURE] (LOW): Audit scripts read local system configuration files to identify potential misconfigurations.
Evidence in scripts/security-checker.sh: Accesses /etc/ssh/sshd_config, /etc/passwd, and /etc/shadow to verify permissions and authentication settings.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill has an attack surface for indirect injection as it processes untrusted local files and external URLs.
Ingestion points: User-provided file paths and URLs processed by vulnerability-scanner.py, headers-validator.sh, and config-auditor.py.
Boundary markers: Absent; data is processed directly.
Capability inventory: File system read access and subprocess command execution are available to the skill.
Sanitization: Absent; the scripts perform regex-based auditing on raw content without prior sanitization.

security-hardening