The Academy

A Socratic dialogue engine for AI agents with Model Context Protocol (MCP) integration. Run multi-agent conversations, execute bulk experiments, and analyze dialogue patterns all controllable through 70+ MCP tools.

Key Features

Feature	Description
Multi-Agent Dialogues	Autonomous conversations between Claude, GPT, Gemini, Grok, and other AI models
Bulk Experiments	Run hundreds of sessions concurrently with configurable parallelism
Live Analysis	Real-time AI-powered conversation analysis with customizable output schema
MCP Integration	70+ tools exposing all functionality programmatically
8 AI Providers	Claude, GPT, Gemini, Grok, Deepseek, Mistral, Cohere, and Ollama
Moderator Controls	Pause, resume, and inject guidance during conversations

Bulk Experiments

Design and execute large-scale conversation studies with the visual experiment designer.

MCP Integration

The Academy exposes an MCP server at /api/mcp with 70+ tools organized into categories:

MCP Resources

Resource	Description
academy://sessions	All conversation sessions with metadata
academy://session/{id}	Individual session with participants and messages
academy://session/{id}/messages	Complete message history
academy://session/{id}/analysis	Analysis snapshots
academy://experiments	All experiment configurations
academy://experiment/{id}/results	Experiment results and analytics

MCP Tools

Session Management

Tool	Description
create_session	Create new conversation session
get_session	Get session by ID (includes analysis snapshots)
get_sessions	List all sessions
delete_session	Delete a session
update_session	Update session metadata
switch_current_session	Change active session
duplicate_session	Clone existing session
import_session	Import session data
list_templates	List session templates
create_session_from_template	Create from template
get_current_session_id	Get active session ID
get_session_analysis_config	Get analysis settings
update_session_analysis_config	Update analysis settings
get_session_chat_config	Get chat settings
update_session_chat_config	Update chat settings

Participant Management

Tool	Description
add_participant	Add AI agent to conversation
remove_participant	Remove participant
update_participant	Modify participant settings
update_participant_status	Change participant state
get_participant_config	Get participant configuration
list_available_models	List available AI models

Conversation Control

Tool	Description
start_conversation	Begin autonomous dialogue
pause_conversation	Pause active conversation
resume_conversation	Resume paused conversation
stop_conversation	End conversation
get_conversation_status	Check conversation state
get_conversation_stats	Get conversation metrics

Message Management

Tool	Description
send_message	Send message to session
update_message	Update message content
delete_message	Delete a message
clear_messages	Clear all messages
inject_moderator_prompt	Insert moderator message

Analysis Tools

Tool	Description
analyze_conversation	Extract insights and patterns
trigger_live_analysis	Run real-time analysis
save_analysis_snapshot	Store analysis data
get_analysis_history	Get past analyses
clear_analysis_history	Remove analysis data
set_analysis_provider	Set analysis AI provider
get_analysis_providers	List analysis providers
auto_analyze_conversation	Toggle auto-analysis

Export Tools

Tool	Description
export_session	Export conversation data (JSON/CSV)
export_analysis_timeline	Export analysis history
get_export_preview	Preview export content

AI Provider Tools

Tool	Description
claude_chat	Claude API with retry logic
openai_chat	OpenAI API with retry logic
grok_chat	Grok API with retry logic
gemini_chat	Gemini API with retry logic
deepseek_chat	Deepseek API with retry logic
mistral_chat	Mistral API with retry logic
cohere_chat	Cohere API with retry logic
ollama_chat	Ollama API with retry logic

Experiment Management

Tool	Description
create_experiment	Create experiment configuration
get_experiments	List all experiments
get_experiment	Get experiment details
update_experiment	Update experiment config
delete_experiment	Delete experiment
create_experiment_run	Create new run
update_experiment_run	Update run status
get_experiment_run	Get run details

Experiment Execution

Tool	Description
execute_experiment	Execute bulk experiment
get_experiment_status	Get execution progress
pause_experiment	Pause running experiment
resume_experiment	Resume paused experiment
stop_experiment	Stop experiment
get_experiment_results	Get aggregated results

Debug Tools

Tool	Description
debug_store	Debug database state
get_api_errors	Get API errors with retry details
clear_api_errors	Clear error logs
log_api_error	Log an API error

Example Usage

// Start a conversation
await mcp.callTool('start_conversation', { sessionId, initialPrompt })

// Run bulk experiment
const exp = await mcp.callTool('create_experiment', {
  config: { name: 'Study', totalSessions: 50, concurrentSessions: 5 }
})
await mcp.callTool('execute_experiment', { experimentId: exp.experimentId })

// Export with analysis
await mcp.callTool('export_session', { sessionId, includeAnalysis: true })

Use Cases

• Multi-agent research — Study AI model interactions in extended conversations
• Parameter studies — Test temperature, prompts, and settings systematically
• Model comparison — Compare AI models on identical tasks
• Bulk experiments — Run large-scale studies programmatically
• Synthetic data generation — Create conversational datasets

Tech Stack

Next.js 15 • PostgreSQL • TypeScript • Tailwind CSS • MCP (JSON-RPC 2.0) • Docker

Getting Started

Prerequisites

• Node.js 18+
• Docker (for PostgreSQL)
• API keys for providers you want to use

Quick Start

# Clone and start database
git clone https://github.com/yourname/the-academy.git
cd the-academy
docker-compose up -d

# Install and run
cd academy
pnpm install
pnpm dev

Environment Variables

Create academy/.env.local:

# AI Providers (add keys for providers you want to use)
ANTHROPIC_API_KEY=your_key
OPENAI_API_KEY=your_key
XAI_API_KEY=your_key
GOOGLE_AI_API_KEY=your_key
DEEPSEEK_API_KEY=your_key
MISTRAL_API_KEY=your_key
COHERE_API_KEY=your_key

# Database
DATABASE_URL=postgresql://academy_user:academy_password@localhost:5432/academy_db

Visit http://localhost:3000

Contributing

Contributions welcome! Areas of interest:

• Additional AI provider integrations
• Analysis algorithms and templates
• UI/UX improvements
• Export format extensions

License

MIT License