Orchestrator Maestro Skill

The Orchestrator Maestro is a meta-skill that coordinates multiple skills into cohesive workflows. It acts as the conductor of the Optimus Pryme skill orchestra, determining execution order, managing dependencies, and preserving context.

Core Capabilities

1. Workflow Choreography

•Determine optimal execution sequence for multi-skill operations
•Parallel execution when dependencies allow
•Sequential execution when required
•Context handoff between skills

2. Dependency Resolution

•Analyze skill requirements and outputs
•Build dependency graphs
•Detect circular dependencies
•Auto-resolve execution order

3. Context Preservation

•Maintain conversation state across skill invocations
•Pass relevant context between skills
•Aggregate results from multiple skills
•Track workflow state

4. Goal Decomposition

•Break complex user requests into skill sequences
•Map intents to appropriate skills
•Generate efficient execution plans
•Adaptive replanning on failures

5. Workflow Templates

•Pre-built workflow patterns for common operations
•Template library management
•Custom workflow creation
•Template versioning

Workflow Templates

Template: New Product Launch

yaml

name: new_product_launch
description: Complete product launch workflow
skills:
  - skill: market-researcher
    input: { asin: "{{product_asin}}", depth: "comprehensive" }
    output_key: market_data
    
  - skill: amazon-listing-optimizer
    input: { asin: "{{product_asin}}", market_insights: "{{market_data}}" }
    output_key: listing_optimizations
    parallel: false
    
  - skill: grok-admaster-operator
    action: create_launch_campaign
    input: { 
      product: "{{product_asin}}",
      strategy: "aggressive",
      keywords: "{{market_data.keywords}}"
    }
    parallel: false

Template: Performance Crisis Response

yaml

name: crisis_response
description: Rapid response to performance issues
skills:
  - skill: grok-admaster-operator
    action: get_dashboard_summary
    output_key: current_metrics
    
  - skill: consciousness-engine
    action: diagnose_issues
    input: { metrics: "{{current_metrics}}" }
    output_key: diagnosis
    parallel: false
    
  - skill: simulation-lab
    action: test_recovery_strategies
    input: { diagnosis: "{{diagnosis}}" }
    output_key: recovery_plans
    parallel: false
    
  - skill: grok-admaster-operator
    action: execute_plan
    input: { plan: "{{recovery_plans.best_plan}}" }
    parallel: false

Template: Quarterly Strategy Refresh

yaml

name: quarterly_strategy_refresh
description: Comprehensive quarterly review and optimization
skills:
  - skill: memory-palace
    action: retrieve_quarterly_patterns
    output_key: historical_insights
    
  - skill: knowledge-synthesizer
    action: analyze_market_trends
    output_key: market_trends
    parallel: true
    
  - skill: competitive-intelligence
    action: competitor_analysis
    output_key: competitive_landscape
    parallel: true
    
  - skill: evolution-engine
    action: evolve_strategies
    input: {
      history: "{{historical_insights}}",
      trends: "{{market_trends}}",
      competition: "{{competitive_landscape}}"
    }
    output_key: new_strategies
    parallel: false

API Operations

Execute Workflow

Endpoint: Internal skill invocation (not HTTP API)

Input:

json

{
  "workflow_template": "new_product_launch",
  "parameters": {
    "product_asin": "B0DWK3C1R7"
  },
  "execution_mode": "sequential|parallel_where_possible",
  "dry_run": false
}

Output:

json

{
  "workflow_id": "wf_abc123",
  "status": "completed",
  "execution_time_seconds": 45.2,
  "skills_executed": 3,
  "results": {
    "market_data": {...},
    "listing_optimizations": {...},
    "campaign_created": {...}
  },
  "execution_log": [
    {
      "skill": "market-researcher",
      "status": "completed",
      "duration_seconds": 12.3
    }
  ]
}

Create Custom Workflow

json

{
  "action": "create_workflow_template",
  "name": "custom_optimization_flow",
  "description": "My custom workflow",
  "skills_sequence": [
    {
      "skill": "grok-admaster-operator",
      "action": "analyze_campaigns"
    },
    {
      "skill": "simulation-lab",
      "action": "test_strategies"
    }
  ]
}

Usage Patterns

Pattern 1: Chain Multiple Skills

Scenario: Research market → Optimize listing → Launch campaign

code

USER: "Launch my new charger product (ASIN B0DWK3C1R7) with optimized listings and PPC"

ORCHESTRATOR:
1. Invoke market-researcher(asin=B0DWK3C1R7)
2. Pass results to amazon-listing-optimizer
3. Create campaign via grok-admaster-operator
4. Return consolidated results

Pattern 2: Parallel Execution

Scenario: Run independent analyses simultaneously

code

USER: "Give me a complete competitive overview"

ORCHESTRATOR:
1. Parallel execution:
   - market-researcher: trend analysis
   - competitive-intelligence: competitor data
   - knowledge-synthesizer: industry insights
2. Aggregate all results
3. Generate unified report

Pattern 3: Adaptive Replanning

Scenario: Skill fails, orchestrator adapts

code

WORKFLOW:
1. Try primary strategy
2. IF FAILS → Fall back to alternative
3. Continue workflow with adjusted plan

Integration with Other Skills

All skills can be orchestrated. The maestro:

•Reads skill SKILL.md files to understand capabilities
•Parses input/output specifications
•Builds dependency chains
•Executes in optimal order

Database Schema

Workflows and executions are tracked in the database:

sql

-- Defined in server/updates/04_meta_skills_tables.sql
skill_executions (
  workflow_id,
  skill_name,
  input_data,
  output_data,
  execution_order,
  status,
  started_at,
  completed_at
)

workflow_templates (
  name,
  description,
  skill_sequence,
  created_at
)

Files

code

.agent/skills/orchestrator-maestro/
├── SKILL.md                          # This file
├── scripts/
│   ├── workflow_engine.py            # Core execution engine
│   ├── dependency_resolver.py        # Dependency graph builder
│   └── context_manager.py            # Context preservation
├── resources/
│   ├── workflow_templates.json       # Template library
│   └── skill_registry.json           # Known skills catalog
└── tests/
    └── test_workflow_engine.py       # Unit tests

Example Invocation

code

USER: "I want to launch a new product. ASIN is B0ABC123. Do the full research, optimize the listing, and create an aggressive PPC campaign."

ORCHESTRATOR ACTION:
1. Detect intent: product launch
2. Load template: new_product_launch
3. Inject parameters: {asin: "B0ABC123"}
4. Execute workflow:
   - market-researcher → 12s
   - amazon-listing-optimizer → 8s
   - grok-admaster-operator (create campaign) → 5s
5. Return: "Product launch workflow completed. Market research shows high demand for your category. Listing optimized with 15 keyword improvements. Campaign 'Aggressive Launch B0ABC123' created with $50/day budget."

Notes

•The orchestrator runs within the agent context, not as a separate service
•It coordinates existing skills but doesn't replace them
•Workflows are reusable and version-controlled
•Failed skills can trigger fallback strategies
•All executions are logged for analysis by consciousness-engine

This skill enables seamless multi-skill operations, turning Optimus Pryme into a true autonomous system.