The Agent Skills Directory

[DATA_EXPOSURE_AND_EXFILTRATION]: The skill demonstrates strong security awareness regarding sensitive data. It explicitly instructs users to set acme.json permissions to mode 600 to protect certificate data and recommends using environment variables or secret mounts for DNS API tokens rather than hardcoding them.
[PRIVILEGE_ESCALATION]: The skill recommends using security_opt: [no-new-privileges:true] in Docker Compose configurations, which is a security best practice to prevent processes from gaining new privileges via setuid or setgid binaries.
[COMMAND_EXECUTION]: Shell commands provided in the documentation (e.g., chmod, docker compose, curl -I) are standard operational procedures for deploying and validating the service and do not exhibit malicious patterns.
[EXTERNAL_DOWNLOADS]: The skill references official Traefik Docker images (traefik:v3.4, traefik/whoami:v1.10) and official documentation from traefik.io, which are well-known and trusted sources in the DevOps ecosystem.
[DYNAMIC_EXECUTION]: A Python script is included in gpt.md to validate Traefik configurations. It uses yaml.safe_load() to parse YAML files, which is the secure way to load untrusted data and prevents arbitrary code execution during parsing.

traefik