The Agent Skills Directory

[SAFE]: The skill instructions are purely technical and focused on performance optimization and data safety. No malicious patterns were detected.
[PROMPT_INJECTION]: The skill defines a clear, professional role for the AI agent. There are no attempts to bypass safety filters or override core instructions in a malicious way. The "Critical Safety Rule" section reinforces safety rather than undermining it.
[COMMAND_EXECUTION]: The skill contains specific guidance to prevent the accidental or unauthorized execution of destructive commands such as DROP TABLE, TRUNCATE, or DELETE. It mandates user confirmation and backups before any such actions are taken.
[DATA_EXFILTRATION]: No patterns for data exfiltration, network requests to untrusted domains, or unauthorized access to sensitive file paths were identified.
[EXTERNAL_DOWNLOADS]: The skill does not reference external scripts, packages, or remote code execution patterns.
[CREDENTIALS_UNSAFE]: The skill does not contain hardcoded credentials and provides sound advice against storing sensitive data in plain text within systems like Redis.

database-optimization