The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides numerous commands to be executed with elevated privileges to manage the system environment, including certbot, snap, apt, systemctl, and tee. These are used to install software, modify configuration files, and manage background services.
[PRIVILEGE_ESCALATION]: The instructions rely heavily on sudo for software installation, certificate issuance, and service management. While these permissions are required for the stated purpose of managing system-level TLS certificates and web server configurations, they grant the agent high-level administrative access to the host.
[INDIRECT_PROMPT_INJECTION]: The skill demonstrates how to interpolate user-provided data, such as domain names and file paths, directly into shell commands (e.g., sudo certbot --nginx -d {domain}). This creates an attack surface for command injection if an agent processes untrusted input without proper sanitization.
Ingestion points: User-supplied domain names and directory paths used in issuance and renewal commands.
Boundary markers: None present in the provided command templates or instructions to delimit user input.
Capability inventory: High-privilege command execution via sudo across multiple utilities (certbot, apt, snap, systemctl, tee).
Sanitization: No sanitization, escaping, or validation logic is defined for the user-supplied variables within the skill.

letsencrypt