The Agent Skills Directory

[SAFE]: The skill contains no malicious patterns, obfuscation, or unauthorized network operations. Its instructions are focused on structured dialogue with the user to reach a shared understanding of a design.
[DATA_EXPOSURE]: The skill instructs the agent to 'explore the codebase' when possible. This is a functional requirement for coding assistants to provide context-aware answers and does not involve data exfiltration or access to sensitive system credentials.
[PROMPT_INJECTION]: The skill uses instructional language to guide the agent's behavior ('Interview me relentlessly', 'Ask the questions one at a time'). These are benign behavioral constraints and do not attempt to bypass safety filters or extract system prompts.

grill-me