Use when the primary agent must decide whether to delegate via the Task tool, which subagent id to call, and how to keep child-session prompts small. Use when multiple subagents could apply, or before large read-only review, verification, security review, or external doc research.
Use when analyzing, reviewing, refactoring, or writing Python code. Apply Pythonic idioms, SOLID design, and Liskov-safe subtyping. Invoke for code review, architecture feedback, or implementation in Python.
Investigate the security posture of a VPS, remote server, or infrastructure. Orchestrates specialized subagents for comprehensive assessment. Use when the user asks to audit, check, harden, scan, or investigate server security, vulnerabilities, compliance, exposed services, SSH hardening, firewall rules, or infrastructure posture. Triggers on VPS, server, security, audit, hardening, pentest, vulnerability, compliance, CIS benchmark, infrastructure security.
Task management CLI for tracking and managing feature subtasks with status, dependencies, and validation