Monomind Monomind Automation Smart Agents

---
description: automation smart-agents command (monomind)
type: flow
name: monomind-monomind-automation-smart-agents
---

# Smart Agent Auto-Spawning

## Purpose
Automatically spawn the right agents at the right time without manual intervention.

## Auto-Spawning Triggers

### 1. File Type Detection
When editing files, agents auto-spawn:
- **JavaScript/TypeScript**: Coder agent
- **Markdown**: Researcher agent
- **JSON/YAML**: Analyst agent
- **Multiple files**: Coordinator agent

### 2. Task Complexity
```
Simple task: "Fix typo"
→ Single coordinator agent

Complex task: "Implement OAuth with Google"
→ Architect + Coder + Tester + Researcher
```

### 3. Dynamic Scaling
The system monitors workload and spawns additional agents when:
- Task queue grows
- Complexity increases
- Parallel opportunities exist

**Status Monitoring:**
```javascript
// Check swarm health
mcp__monomind__swarm_status({
  swarmId: "current"
})

// Monitor agent status and health
mcp__monomind__agent_status({
  id: "agent-123"
})
mcp__monomind__agent_health({
  id: "agent-123",
  detailed: true
})
```

## Configuration

### MCP Tool Integration
Uses Monomind MCP tools for agent coordination:
```javascript
// Initialize swarm with appropriate topology
mcp__monomind__swarm_init({
  "topology": "mesh",
  "maxAgents": 8,
  "strategy": "auto"
})

// Spawn agents based on file type
mcp__monomind__agent_spawn({
  "type": "coder",
  "name": "JavaScript Handler",
  "capabilities": ["javascript", "typescript"]
})
```

### Fallback Configuration
If MCP tools are unavailable:
```bash
npx monomind hooks pre-task --description "<task>" --auto-spawn
```

## Benefits
- Zero manual agent management
- Perfect agent selection
- Dynamic scaling
- Resource efficiency