The Agent Skills Directory

[SAFE]: The skill consists of plain-text instructions that guide the agent's conversational persona. It does not include any executable scripts or configuration that could be used maliciously.
[PROMPT_INJECTION]: No attempts to bypass safety filters or override system instructions were found. The 'IMPORTANT' keyword is not used in a malicious context.
[DATA_EXPOSURE]: The skill encourages the agent to explore the local codebase to answer questions. This is a standard and expected behavior for development-oriented agents and does not involve exfiltration or access to sensitive credentials.

grill-me