The Agent Skills Directory

[SAFE]: The skill consists entirely of instructional text designed to improve the quality and completeness of the agent's output. It explicitly bans common placeholders like // ... and TODO to ensure full code generation.- [SAFE]: There are no indicators of prompt injection, data exfiltration, persistence mechanisms, or unauthorized network activity. The instructions are focused on formatting and output management.- [SAFE]: No code, scripts, or external dependencies are included in the skill, precluding risks associated with remote code execution or privilege escalation.

full-output-enforcement