The Agent Skills Directory

[SAFE]: No security issues or malicious patterns were identified during the analysis.\n- [PROMPT_INJECTION]: The skill uses persona-based instructions to guide the AI's perspective. These instructions are standard for role-definition and do not contain commands to bypass safety filters, override system constraints, or ignore previous instructions.\n- [DATA_EXFILTRATION]: The skill does not perform any network operations, access sensitive file paths, or attempt to exfiltrate user data. There are no hardcoded credentials or unauthorized data access patterns.\n- [COMMAND_EXECUTION]: There are no executable scripts, shell commands, or platform-specific dynamic context injections present in the skill files. It consists entirely of instructional markdown and metadata.

ford-perspective