The Agent Skills Directory

[SAFE]: The skill demonstrates a strong security posture by instructing the agent to generate all secrets (encryption keys, database passwords) locally on the target machine using cryptographically secure methods, and ensuring they are stored in files with restricted permissions (chmod 600).
[REMOTE_CODE_EXECUTION]: The deployment flow includes installing Docker via the official script at get.docker.com. Following the Trusted Source Rule, this reference is documented as safe for the intended administrative purpose.
[COMMAND_EXECUTION]: The skill provides the agent with necessary administrative commands to provision the server, including managing the ufw firewall and orchestrating Docker containers. These commands are scoped to the deployment task and adhere to the principle of least privilege where possible (e.g., isolating internal services from the public internet).
[DATA_EXFILTRATION]: A network request to ifconfig.me is used to determine the server's public IP address for DNS preflight checks. This is a standard operation for automated SSL certificate issuance and does not involve unauthorized data transmission.
[INDIRECT_PROMPT_INJECTION]: The skill defines clear boundaries for user-supplied data (such as domain names and timezones) which are used to configure the environment. The risk of injection is mitigated by the agent's structured execution of the deployment steps and the use of local generation for sensitive values.

n8n-self-hosting