AI Orchestrator
Integrates with Google's Gemini API to power AI orchestration strategies including councils, debates, and evaluations.
Integrates with local Ollama models to run AI orchestration strategies without external API keys.
Integrates with OpenAI's API to power AI orchestration strategies including councils, debates, and evaluations.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@AI OrchestratorCouncil review: Should we adopt microservices?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
🤖 AI Orchestrator
Advanced AI orchestration engine — multi-advisor councils, structured debates, creative brainstorming, and multi-criteria evaluation. Works as an MCP (Model Context Protocol) server with any compatible AI tool.
Supports: OpenCode, Claude Code, Cursor, Windsurf, Continue, Cline, Copilot, and any MCP-compatible client.
✨ Features
5 Orchestration Strategies
Strategy | Description | Use Case |
Council 🏛️ | Multi-advisor decision review with anonymous peer evaluation (Karpathy method) | Critical decisions, risk assessment |
Debate ⚔️ | Structured pro/con debate with judge verdict | Go/no-go decisions, tradeoff analysis |
Brainstorm 💡 | Creative idea generation with scoring and clustering | Feature discovery, product innovation |
Evaluate 📊 | Multi-criteria option scoring with weighted analysis | Vendor selection, tech choices |
Spec Review 🔍 | Specialist review from multiple angles (security, UX, DevOps, etc.) | Pre-implementation plan review |
15+ Advisor Personas
Persona | Stance | Specialty | Language |
Skeptic (Muhalif) | Critical | Risk detection, failure modes | EN / TR |
First Principles (İlk İlkeler) | Analytical | Assumption deconstruction | EN / TR |
Expansionist (Genişlemeci) | Creative | Opportunity discovery | EN / TR |
Outsider (Yabancı) | Neutral | Jargon-free fresh eyes | EN / TR |
Executor (İcracı) | Practical | Implementation, first steps | EN / TR |
Visionary | Creative | Future opportunities | |
Pragmatist | Practical | Feasibility, resource constraints | |
Security Auditor | Critical | Attack surfaces, data exposure | |
UX Advocate | Critical | User journey, accessibility | |
Business Analyst | Analytical | ROI, market positioning | |
DevOps Engineer | Practical | Deployability, scaling | |
Ethicist | Analytical | Fairness, societal impact | |
Architect | Analytical | System design, patterns | |
Tester | Critical | Edge cases, testability | |
Growth Hacker | Creative | Metrics, viral loops |
Turkish advisors (Muhalif, İlk İlkeler, Genişlemeci, Yabancı, İcracı) are available for Turkish-language decisions. The rest respond in English by default but support any language the question is asked in.
Pre-built Councils
Council | Advisors | Language | Use Case |
Executive Board | Skeptic, Visionary, Pragmatist, Security, Business, UX | EN | General business decisions |
Akıl Kurulu | Muhalif, İlk İlkeler, Genişlemeci, Yabancı, İcracı | TR | Turkish-language decisions |
Tech Review | Architect, Security, DevOps, Tester | EN | Architecture/stability review |
Ethics Board | Ethicist, UX, Skeptic, Business | EN | Ethical impact assessment |
Quick Check | Skeptic, Pragmatist, Visionary | EN | Fast 3-advisor decision |
Multi-Provider Support
Provider | Setup |
OpenAI |
|
Anthropic (Claude) |
|
OpenRouter |
|
Ollama (local) | No key needed — just run |
Groq |
|
Google (Gemini) |
|
DeepSeek |
|
Related MCP server: Claude Team MCP
🚀 Quick Start
Prerequisites
Node.js 18+
At least one LLM provider API key (or Ollama running locally)
Install & Configure
# Clone
git clone https://github.com/YOUR_USER/ai-orchestrator.git
cd ai-orchestrator
# Install
npm install
npm run build
# Configure at least one provider
export OPENAI_API_KEY="sk-..."
# OR
export ANTHROPIC_API_KEY="sk-ant-..."
# OR run Ollama locally: ollama serveAdd to Your AI Tool
OpenCode
// opencode.json
{
"mcpServers": {
"ai-orchestrator": {
"command": "node",
"args": ["/path/to/ai-orchestrator/dist/index.js"],
"env": {
"OPENAI_API_KEY": "${OPENAI_API_KEY}"
}
}
}
}Claude Code
// .claude/mcp.json
{
"mcpServers": {
"ai-orchestrator": {
"command": "node",
"args": ["/path/to/ai-orchestrator/dist/index.js"],
"env": {
"ANTHROPIC_API_KEY": "${ANTHROPIC_API_KEY}"
}
}
}
}Cursor / Windsurf
Add to Cursor MCP settings or .cursor/mcp.json:
{
"mcpServers": {
"ai-orchestrator": {
"command": "node",
"args": ["/absolute/path/to/ai-orchestrator/dist/index.js"]
}
}
}📖 Usage
Once configured, call any tool from your AI assistant:
Council (Decision Review)
Use the orchestrate tool to review: "Should we rewrite our Bootstrap themes to Tailwind?"
Strategy: council
Council: executive_boardReturns: 6 independent advisor opinions → anonymous peer review → chairman synthesis with recommendations.
Debate (Pro/Con Analysis)
Orchestrate a debate on: "Should we adopt microservices?"
Strategy: debate
Rounds: 3Brainstorm (Idea Generation)
Brainstorm ideas for: "How to improve developer onboarding?"
Strategy: brainstormEvaluate (Option Scoring)
Evaluate these database options: PostgreSQL, MongoDB, Supabase
Criteria: Performance (0.4), Ease of Use (0.3), Cost (0.3)
Strategy: evaluateSpec Review (Plan Audit)
Review this architecture spec from security, performance, and UX perspectives.
Strategy: spec-reviewCouncil with Per-Advisor Provider Routing
Each advisor can use a different LLM provider/model when multiple API keys are configured:
Evaluate: "Should we migrate to microservices?"
Strategy: council
Council: executive_board
Advisors:
- id: skeptic
provider: deepseek
- id: visionary
provider: anthropic
- id: pragmatist
provider: openai
model: gpt-4o
- id: security_auditor
provider: groq
- id: business_analyst
provider: google
- id: ux_advocate
provider: openrouterAdvisors without a provider spec use auto-detected providers in round-robin. If no API keys are configured, simulation mode runs all advisors using the calling AI's own intelligence.
🔧 Architecture
src/
├── index.ts # MCP Server entry point (4 tools exposed)
├── types.ts # TypeScript type definitions
├── personas.ts # 15+ advisor personas + 5 council presets
├── providers.ts # LLM provider layer (7 providers)
└── strategies/
├── council.ts # Karpathy 3-stage multi-advisor council
├── debate.ts # Structured pro/con debate with judge
├── brainstorm.ts # Creative brainstorming with clustering
└── evaluate.ts # Multi-criteria evaluation + spec reviewCouncil Flow (Karpathy Method)
┌─────────────┐
│ Question │
└──────┬──────┘
┌──────────────┼──────────────┐
┌─────▼─────┐ ┌─────▼─────┐ ┌─────▼─────┐
│ Advisor A │ │ Advisor B │ │ Advisor N │ ← Parallel (isolated)
│ (Skeptic) │ │(Visionary)│ │ (...) │
└─────┬─────┘ └─────┬─────┘ └─────┬─────┘
└──────────────┼──────────────┘
┌─────▼─────┐
│ Anonymize │ ← Anonymous review
└─────┬─────┘
┌──────────────┼──────────────┐
┌─────▼─────┐ ┌─────▼─────┐ ┌─────▼─────┐
│ Peer Rev A│ │ Peer Rev B│ │ Peer Rev N│
└─────┬─────┘ └─────┬─────┘ └─────┬─────┘
└──────────────┼──────────────┘
┌─────▼─────┐
│ Chairman │ ← Synthesis
│ Verdict │
└───────────┘🌐 Multi-Language Support
The system is English-first but fully multi-language:
Executive Board, Tech Review, Ethics Board, Quick Check → English by default
Akıl Kurulu → Turkish advisors for Turkish-language decisions
You can ask questions in any language — the system matches the input language
Advisors defined in Turkish (Muhalif, İlk İlkeler, etc.) will respond in Turkish when used
The simulation mode (no API keys) also supports any language the calling AI understands
🛠️ Development
npm install # Install dependencies
npm run build # Compile TypeScript
npm run dev # Watch mode
npm start # Run server directly📄 License
MIT — see LICENSE
🙏 Credits
Inspired by Andrej Karpathy's LLM Council
Built on the Model Context Protocol
Advisor personas adapted from the Claude Skills LLM Council community
Made for AI-powered decision making. Ship with confidence.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables AI assistants to create council hearings where multiple LLMs (Claude, GPT, Gemini, Grok) debate topics and synthesize verdicts with trust scores and diverse perspectives.Last updated230Apache 2.0
- AlicenseBqualityCmaintenanceEnables collaboration between multiple AI models (GPT, Claude, Gemini) to work together on complex tasks, with intelligent task distribution and role-based expert assignment for code development, review, and optimization.Last updated181949MIT
- AlicenseBqualityDmaintenanceEnables users to leverage a team of specialized AI agent personas like Business Analysts and Product Managers through a unified interface across multiple LLM providers. It streamlines agile development workflows by providing tools for model discovery, unified prompt delivery, and role-specific decision making.Last updated9MIT
- Alicense-qualityDmaintenanceEnables Claude to orchestrate tasks across 27 AI providers, run multi-agent plans, and conduct multi-model councils for decision-making.Last updated99MIT
Related MCP Connectors
Build, validate, and deploy multi-agent AI solutions from any AI environment.
Cross-agent artifact workspace with provenance across Claude Code, Codex, Cursor, LangGraph.
The project brain for AI coding agents — memory, decisions, sprints, knowledge base via MCP.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/halitartuc/ai-orchestrator'
If you have feedback or need assistance with the MCP directory API, please join our Discord server