The Agent Skills Directory

[SAFE]: The skill consists of high-level instructions that guide an AI to perform code reviews and verify plan alignment. The instructions explicitly direct the AI to look for security vulnerabilities and maintain code quality, which is a positive safety behavior. No malicious patterns such as prompt injection, persistence mechanisms, or credential harvesting were detected.- [NO_CODE]: There are no scripts, executables, or tool definitions included in this skill. It operates entirely as a set of natural language instructions, which prevents the introduction of traditional code-based attack vectors like command injection or buffer overflows.

code-reviewer