The Agent Skills Directory

[Privilege Escalation] (MEDIUM): The skill frequently employs sudo to perform high-privilege operations, such as installing packages via apt, managing system services with systemctl, and writing to protected directories like /etc/nginx/. While these actions are essential for the primary purpose of server configuration, they provide a broad attack surface if the agent is exploited. Finding evidence in SKILL.md and scripts/ssl-setup.sh.
[External Downloads] (LOW): The skill automates the installation of certbot and its Nginx plugins using system package managers (apt, dnf). Per [TRUST-SCOPE-RULE], these downloads from official operating system repositories are considered low risk. Finding evidence in scripts/ssl-setup.sh (line 14) and references/ssl.md.
[Data Exposure] (LOW): The scripts and documentation reference sensitive filesystem paths, specifically SSL private keys located at /etc/letsencrypt/live/example.com/privkey.pem. While necessary for server configuration, access to these paths should be strictly monitored. No exfiltration behavior was detected.
[Indirect Prompt Injection] (LOW): The skill contains an ingestion surface for untrusted data which is subsequently used in file-system and command operations.
Ingestion points: Positional parameters $domain and $port in scripts/nginx-generator.sh and scripts/ssl-setup.sh.
Boundary markers: Absent; user inputs are interpolated into configuration templates and shell commands without delimiters or warnings to ignore embedded instructions.
Capability inventory: File writing via cat, package installation via apt, and configuration testing via nginx -t.
Sanitization: Absent; the scripts do not perform validation or escaping of the domain and port inputs before they are written to the filesystem.

nginx