Skip to main content

​
What Are MCP Servers?

MCP (Model Context Protocol) servers expose tools that AI models can call to access external data and services. Instead of relying on training data alone, models can fetch current information, interact with APIs, and perform real-world actions. This dramatically reduces hallucinations and enables AI assistants to provide accurate, up-to-date, and contextually grounded responses.

​
How MCP Servers Work

1

AI Needs Information

Your AI model encounters a question that requires external data (e.g., “What’s the weather in Paris?”)
2

Model Identifies Tool

The AI decides which MCP server tool can answer the question (e.g., get-weather from a Weather MCP server)
3

Tool Executes

The tool calls the external API with the necessary parameters
4

Model Responds

The AI incorporates the tool’s response into its final answer to the user

​
Available MCP Servers

Google Maps

Geocoding, place search, directions, and location services

OpenRouter

AI model recommendations and capability information

Time Utilities

Current time, date utilities, and timezone conversion

Reminders

Create, manage, and trigger reminders for conversations

Conversation Archives

Access archived conversation message summaries

​
Authentication Patterns

MCP servers use three authentication patterns:

​
Public (No Authentication)

Some servers are publicly accessible and require no authentication. Examples: Time Utilities, OpenRouter 
# No authentication required
curl https://app.chatbotplatform.io/mcp/time/call/today

​
Conversation API Key

Servers that access conversation-specific data require the conversation’s API key. The key is automatically provided when tools are called within an Agentic Loop. Examples: Reminders, Conversation Archives 
# Requires conversation API key
curl https://app.chatbotplatform.io/mcp/reminders/call/list-reminders \
  -H "Authorization: Bearer conv_xxxxx..."

​
External API Key

Some servers proxy external services and require an API key from that service. Pass the key in a custom header. Examples: Google Maps 
# Requires Google Maps API key
curl https://app.chatbotplatform.io/mcp/google-maps/call/geocode \
  -H "X-Google-Maps-Api-Key: YOUR_API_KEY"

​
Using MCP Servers with Agentic Loops

MCP servers integrate seamlessly with Agentic Loops. When you create an Agentic Loop and configure MCP servers, the AI model automatically has access to all tools provided by those servers.
See Agentic Loops → Using MCP Servers for detailed setup instructions.

​
Building Custom MCP Servers

Chatbot Platform is built on Laravel MCP, making it easy to create your own MCP servers.

Build Your Own MCP Server

Learn how to build custom MCP servers using Laravel’s MCP package

​
When to Use MCP Servers

Use MCP servers when your AI needs to:
  • Fetch current information - Weather, stock prices, real-time data
  • Search external databases - Customer records, product catalogs, documentation
  • Perform actions - Create reminders, send emails, update records
  • Access user-specific data - Account details, order history, preferences
  • Validate information - Check availability, verify addresses, confirm data
MCP servers turn AI from a static text generator into a dynamic assistant that can interact with the real world.

API Reference

View the complete API specification with all MCP server endpoints