Status

Status

Display a compact system status dashboard for the orchestrator.

Usage

• /status - Show full status dashboard

Algorithm

1. Gather data (in parallel where possible):

   Agents: Glob ~/.claude/agents/**/*.md and count by tier:

   • Core (L0): files directly in agents/
   • Expert (L1): files in agents/experts/
   • Specialist (L2): files in agents/specialists/

   Skills: Glob ~/.claude/skills/*/SKILL.md and count

   Rules: Glob ~/.claude/rules/*.md and list names

   Memory: Read ~/.claude/projects/*/memory/MEMORY.md - check if exists, get line count

   Learnings: Read ~/.claude/learnings/instincts.json if exists:

   • Count total instincts
   • Count high-confidence (confidence >= 0.8)

   MCP Plugins: Count available MCP tool prefixes (mcp__*) from tool list

   Sessions: Glob ~/.claude/sessions/**/*.md and count

2. Display dashboard:

   ============================================
    ORCHESTRATOR V11.0 - System Status
   ============================================
   
    Agents
      Core (L0):       6
      Expert (L1):     19
      Specialist (L2): 18
      Total:           43
   
    Skills
      Loaded:          {count}
      User-invocable:  {count}
   
    Rules
      Active sets:     {list}
   
    Memory
      Status:          Loaded
      Size:            {lines} lines
   
    Learnings
      Total instincts: {count}
      High confidence: {count}
   
    MCP Plugins
      Connected:       {count}
   
    Sessions
      Saved:           {count}
   
   ============================================
   

3. Health indicators (append if issues detected):

   • WARN if no memory file found
   • WARN if 0 agents detected
   • WARN if 0 skills detected
   • OK if everything normal

Notes

• This is a read-only status check; it modifies nothing
• Data is gathered from filesystem, not runtime state
• Agent/skill counts reflect what is on disk, not what is loaded in memory