The Agent Skills Directory

[SAFE]: The skill serves as a structural framework for multi-agent coordination. It uses markdown-based instructions to define roles and task flows without executing external code or scripts.
[SAFE]: The skill configuration limits access to standard read-only tools (Read, Glob, Grep), ensuring that the orchestration patterns focus on codebase analysis and discovery without the risk of unauthorized file system modifications or data exfiltration.
[SAFE]: No evidence of prompt injection, obfuscation, or persistence mechanisms was found. The instructions align with documented platform capabilities for agent management.

parallel-agents