Skip to main content

MCP (Model Context Protocol) Introduction

· 3 min read
Eduardez
Eduardez
MoLOS Lead Developer

We're thrilled to announce full support for the Model Context Protocol (MCP) in MoLOS! This protocol enables seamless integration between AI models and external systems, opening up powerful possibilities for automation and data access.

What is MCP?​

The Model Context Protocol is a standardized protocol that allows AI models to interact with external systems through resources and prompts. Think of it as a universal API that enables AI to read, write, and interact with your MoLOS modules in a controlled and secure manner.

Core Concepts​

Resources​

Resources are representations of external data that can be read by AI models. In MoLOS, resources can represent:

  • Module configurations
  • Database records
  • API endpoints
  • File system data

Each resource has a URI and provides content through standardized read operations.

Prompts​

Prompts are templates that provide context and instructions to AI models. They can include:

  • Task descriptions
  • Data formatting instructions
  • Behavioral guidelines
  • Output structure specifications

API Keys​

API keys control access to MCP endpoints, ensuring only authorized clients can interact with your MoLOS instance. Keys can be scoped to specific modules for fine-grained access control.

Key Features​

Secure API Key Management​

Generate and manage API keys through the MCP dashboard. Each key can be:

  • Scoped to specific modules
  • Configured with expiration dates
  • Revoked instantly when no longer needed

Module Integration​

All MoLOS modules can be exposed as MCP resources, allowing AI models to:

  • Read module configurations
  • Query module data
  • Execute module operations (with proper authorization)

OAuth Integration​

MCP endpoints can be protected using OAuth 2.0, enabling third-party applications to interact with your MoLOS instance through secure, delegated access.

Getting Started​

1. Create an API Key​

Navigate to Settings → AI → MCP → API Keys and click "Create API Key":

  • Choose a descriptive name
  • Select which modules the key can access
  • Set an expiration date (optional)
  • Save and copy the generated key

2. Define Resources​

Create MCP resources that represent your data:

  • Choose a resource type (Database, API, File)
  • Configure connection parameters
  • Define the URI structure
  • Set access permissions

3. Configure Prompts​

Create prompts that guide AI interactions:

  • Define the prompt template
  • Specify required parameters
  • Set default values
  • Test the prompt execution

Use Cases​

  • AI-Powered Analytics: Let AI analyze your module data
  • Automated Reporting: Generate reports using prompt templates
  • Cross-Module Operations: Enable AI to coordinate across modules
  • Third-Party Integrations: Provide controlled access to external AI services

Security & Privacy​

  • API keys are stored encrypted at rest
  • OAuth 2.0 support enables secure third-party access
  • Module scoping prevents unauthorized data access
  • Comprehensive audit logging tracks all MCP interactions

View documentation

What's Next​

We're actively working on:

  • Enhanced resource types including streaming responses
  • Prompt templates with dynamic parameter injection
  • MCP client SDKs for popular programming languages
  • Performance optimizations for high-throughput scenarios

Want to dive into the technical implementation? Check out our dev blog series on MCP implementation starting with Devlog Feb 1: MCP Implementation (Part 1).