The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes sudo for administrative actions including service reloads via systemctl, configuration validation, and interaction with the HAProxy runtime API socket using socat. These are standard operations for the intended task.
[COMMAND_EXECUTION]: Instructions are provided for creating SSL certificate bundles and managing file permissions for sensitive PEM files.
[EXTERNAL_DOWNLOADS]: The skill references official documentation and resources from haproxy.org and mozilla.org. Both are recognized as trusted or well-known services.
[PROMPT_INJECTION]: The skill provides an attack surface for indirect prompt injection as it ingests untrusted configuration data.
Ingestion points: Processes files matching **/haproxy.cfg and other configuration patterns in the SKILL.md globs.
Boundary markers: The skill does not implement specific delimiters or warnings to ignore instructions embedded within configuration files.
Capability inventory: The skill can execute privileged shell commands, modify service states, and manipulate load balancer server configurations.
Sanitization: Recommended use of haproxy -c for syntax validation acts as a partial integrity check, but does not sanitize against instruction injection.

haproxy