The Agent Skills Directory

[SAFE]: The skill implements a structured workflow for designing technical specifications. It performs local file operations such as reading project context (e.g., CLAUDE.md) and existing specs to maintain project consistency, and writes the final output to a markdown file in a specs/ directory. No malicious patterns such as data exfiltration, privilege escalation, or remote code execution were detected.
[SAFE]: The skill includes explicit 'Hard rules' that prohibit the AI agent from writing executable code, focusing its capabilities entirely on documentation and requirement clarification.
[SAFE]: No obfuscation, prompt injection attempts, or suspicious dependencies were found in the provided files.

spec