MCP Endpoints

Complete guide to all Model Context Protocol (MCP) endpoints, configuration, and authentication.

Overview
A powerfull meta integration, let your Agents interact with Dialetica trough MCP. All MCP endpoints use the streamable-http transport protocol and require Bearer token authentication.

Base Configuration:

  • Transport: streamable-http
  • Mount Path: /mcp
  • Base URL Pattern: {base_url}/v1/{resource}/mcp
  • Authentication: Bearer token in Authorization header

Example Endpoint URL:

https://api.dialetica.ai/v1/agents/mcp
Authentication
All MCP endpoints require Bearer token authentication.

Include your API key in the Authorization header:

Authorization: Bearer dai_your_api_key_here

Getting Your API Key:

  • 1. Log in to your dashboard
  • 2. Navigate to the API Keys page
  • 3. Click "Create New API Key"
  • 4. Copy the key and store it securely

Agents MCP

Manage AI agents via MCP. Endpoint: /v1/agents/mcp

Configuration
Configure your MCP client to connect to the Agents endpoint

MCP Configuration

{
  "url": "https://api.dialetica.ai/v1/agents/mcp",
  "type": "http",
  "headers": {
    "Authorization": "Bearer dai_your_api_key_here"
  }
}

Available Tools:

  • create_agent - Create a new agent
  • bulk_create_agents - Create multiple agents at once
  • get_agents - Get all agents created by the authenticated user
  • get_agent - Get an agent by ID
  • update_agent - Update an agent
  • delete_agent - Delete an agent

Contexts MCP

Manage conversation contexts and run conversations via MCP. Endpoint: /v1/contexts/mcp

Configuration
Configure your MCP client to connect to the Contexts endpoint

MCP Configuration

{
  "url": "https://api.dialetica.ai/v1/contexts/mcp",
  "type": "http",
  "headers": {
    "Authorization": "Bearer dai_your_api_key_here"
  }
}

Available Tools:

  • create_context - Create a new context
  • bulk_create_contexts - Create multiple contexts at once
  • get_contexts - Get all contexts created by the authenticated user
  • get_context - Get a context by ID
  • update_context - Update a context
  • delete_context - Delete a context
  • run_context - Run a context with messages
  • run_context_stream - Run a context with messages and stream the response
  • subscribe_context - Subscribe to real-time updates for a context
  • cancel_session - Cancel an active session for a context
  • list_sessions - List active sessions for a context
  • get_conversation_history - Get conversation history for a context
  • route_messages - Route messages in a context to determine the next speaker

Crons MCP

Manage scheduled cron jobs via MCP. Endpoint: /v1/crons/mcp

Configuration
Configure your MCP client to connect to the Crons endpoint

MCP Configuration

{
  "url": "https://api.dialetica.ai/v1/crons/mcp",
  "type": "http",
  "headers": {
    "Authorization": "Bearer dai_your_api_key_here"
  }
}

Available Tools:

  • create_cron - Create a new cron job
  • get_crons - Get all crons for the authenticated user
  • get_cron - Get a cron by ID
  • update_cron - Update a cron job
  • delete_cron - Delete a cron job
  • get_crons_by_context - Get all crons for a specific context

Knowledge MCP

Manage knowledge base entries and perform semantic search via MCP. Endpoint: /v1/knowledge/mcp

Configuration
Configure your MCP client to connect to the Knowledge endpoint

MCP Configuration

{
  "url": "https://api.dialetica.ai/v1/knowledge/mcp",
  "type": "http",
  "headers": {
    "Authorization": "Bearer dai_your_api_key_here"
  }
}

Available Tools:

  • get_all_knowledge - Get all knowledge entries created by the authenticated user
  • create_knowledge - Create or update knowledge entry using upsert
  • bulk_create_knowledge - Create multiple knowledge entries at once
  • update_knowledge - Update existing knowledge entry with ownership validation
  • get_knowledge - Get knowledge entry by ID with ownership validation
  • get_knowledge_by_context - Get all knowledge entries for a specific context
  • get_knowledge_by_agent_and_context - Get knowledge entries for a specific agent within a specific context
  • query_knowledge_post - Query knowledge using semantic search (POST version)
  • query_knowledge_by_context - Query knowledge in a context using semantic search
  • delete_knowledge - Delete knowledge

Models MCP

Get information about available AI models via MCP. Endpoint: /v1/agents/mcp

Configuration
Configure your MCP client to connect to the Models endpoint

MCP Configuration

{
  "url": "https://api.dialetica.ai/v1/agents/mcp",
  "type": "http",
  "headers": {
    "Authorization": "Bearer dai_your_api_key_here"
  }
}

Available Tools:

  • get_models - Get list of available AI models with their metadata

Tool Configs MCP

Manage MCP tool configurations via MCP. Endpoint: /v1/tool-configs/mcp

Configuration
Configure your MCP client to connect to the Tool Configs endpoint

MCP Configuration

{
  "url": "https://api.dialetica.ai/v1/tool-configs/mcp",
  "type": "http",
  "headers": {
    "Authorization": "Bearer dai_your_api_key_here"
  }
}

Available Tools:

  • create_tool_config - Create a new tool configuration
  • get_tool_configs - Get all tool configs for the authenticated user
  • get_tool_config - Get a specific tool config by ID
  • update_tool_config - Update a tool config
  • delete_tool_config - Delete a tool config

Usage MCP

Get usage statistics and tracking information via MCP. Endpoint: /v1/usage/mcp

Configuration
Configure your MCP client to connect to the Usage endpoint

MCP Configuration

{
  "url": "https://api.dialetica.ai/v1/usage/mcp",
  "type": "http",
  "headers": {
    "Authorization": "Bearer dai_your_api_key_here"
  }
}

Available Tools:

  • get_usage_summary - Get usage summary for the authenticated user
  • get_usage - Get detailed usage records for the authenticated user
Transport Protocol Details
Understanding the streamable-http transport

Streamable HTTP Transport:

The streamable-http transport is used for all MCP endpoints. This transport:

  • • Uses HTTP/HTTPS for communication
  • • Supports streaming responses for long-running operations
  • • Requires Bearer token authentication in the Authorization header
  • • Supports standard MCP protocol messages (initialize, tools/list, tools/call, etc.)

Connection Parameters:

  • url: The full endpoint URL
  • headers: HTTP headers including the Authorization header
  • timeout: Request timeout in seconds (default: 30)
  • max_retry_attempts: Number of retry attempts (default: 3)
  • client_session_timeout_seconds: Client session timeout (default: 60)
Best Practices
Tips for using MCP endpoints effectively

Security:

  • • Store API keys securely
  • • Never commit API keys to version control
  • • Use environment variables
  • • Rotate keys regularly

Performance:

  • • Reuse MCP client connections
  • • Implement proper error handling
  • • Monitor usage via Usage MCP
  • • Handle rate limits gracefully
Ready to Start Using MCP?
Explore our guides and examples to learn more about building with Dialetica AI MCP endpoints.