The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes multiple shell commands to interact with the system environment and Nginx service. It specifically uses sudo (via ${NGINX_SUDO:+sudo}) to run commands such as nginx -t, nginx -T, ss, and netstat. While necessary for Nginx management, this grants the agent elevated privileges on the host system.
[DATA_EXFILTRATION]: The skill accesses sensitive system resources, specifically Nginx access logs, error logs, and SSL certificate files (e.g., /var/log/nginx/*.log). This metadata can contain sensitive information about web traffic, client identities, and security configurations.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the processing of untrusted log data.
Ingestion points: Nginx logs are read using tail and awk commands in SKILL.md.
Boundary markers: The skill does not implement boundary markers or instructions to ignore embedded commands within the ingested logs.
Capability inventory: The skill possesses significant capabilities, including shell execution and administrative access via sudo.
Sanitization: Log content is not sanitized or escaped before being processed by the agent.

managing-nginx