The Agent Skills Directory

[SAFE]: The skill is a management utility for repository organization, ensuring consistent instructions across multiple AI agents (Claude, Gemini, Copilot, etc.).
[SAFE]: All operations are performed locally using standard shell commands (ln, mkdir, mv). No external network requests, data exfiltration, or remote downloads were detected.
[SAFE]: The scripts include explicit safety checks to prevent data loss, such as refusing to overwrite existing files unless the --force flag is provided, in which case it creates .bak backups.
[SAFE]: The tool specifically identifies and excludes sensitive configuration files (MCP configurations) from its synchronization logic, advising the user that these files may contain API keys and should be handled manually.
[SAFE]: No obfuscation, prompt injection, or privilege escalation patterns were found in the instructions or the accompanying scripts.

agent-rules