✨ Give Your AI New Skills & Abilities

SMCP makes it easy to enhance your conversational and coding AI tools with powerful plugins. Vibe code your own plugins, extend your AI's capabilities, and connect it to the tools and services you use every day. From productivity and content creation to communications and even roboticsβ€”SMCP gives your AI the skills it needs to do more.

Why Do You Need SMCP?

Your AI assistant is powerful, but it can only do what it's trained to do. SMCP bridges the gap between your AI and the real world, giving it access to tools, services, and capabilities it never had beforeβ€”without retraining or complex integrations.

🎯 Give Your AI Real Skills

Connect your AI to calendars, email, APIs, databases, and more. Each plugin is a new skill your AI can use to help you accomplish real tasks.

⚑ Easy Plugin Creation

Vibe code your own plugins in minutes. No complex frameworks or deep technical knowledge requiredβ€”just simple Python scripts that extend your AI's abilities.

πŸ”Œ Works with Any AI Tool

Built on the Model Context Protocol standard, SMCP works with popular AI frameworks and tools. Originally built for Letta, now compatible with the entire MCP ecosystem.

What Can Your AI Do with SMCP?

The possibilities are endless. Here are just a few ways SMCP can enhance your AI's capabilities:

πŸ“Š Productivity

Manage your calendar, send emails, create tasks, access databases, and automate workflows. Your AI becomes your personal productivity assistant.

🎨 Content Creation

Generate images, edit videos, publish content, manage social media, and create multimedia projects. Turn your AI into a creative powerhouse.

πŸ’¬ Communications

Send messages on Slack, Discord, Telegram, manage customer support, and automate communications across all your platforms.

πŸ€– Robotics & Embodied AI

Control hardware, read sensors, manage actuators, and build physical AI systems. SMCP brings your AI into the real world.

Why SMCP is Different

Built on Open Standards

SMCP implements the Model Context Protocol (MCP) specification, ensuring compatibility with the entire MCP ecosystem. For developers: we use JSON-RPC 2.0 for message handling and Server-Sent Events (SSE) for real-time communication.

Endpoints

SSE Endpoint

GET /sse

Establishes a persistent Server-Sent Events connection for real-time server-to-client communication.

Message Endpoint

POST /messages/

Handles JSON-RPC 2.0 requests from clients for tool execution and protocol operations.

Message Format

All messages follow the JSON-RPC 2.0 specification:

{
  "jsonrpc": "2.0",
  "id": "unique-request-id",
  "method": "method-name",
  "params": {
    // Method-specific parameters
  }
}

Core MCP Methods

Method Description
initialize Initializes the MCP connection and negotiates protocol version
tools/list Lists all available tools from registered plugins
tools/call Executes a tool with provided arguments
health Returns server health status and metrics

How It Works

SMCP uses a simple plugin system that automatically discovers and activates new skills. Just add a plugin, and your AI gains new abilities instantly.

Plugin Discovery

The server automatically discovers plugins by scanning the plugin directory for subdirectories containing a cli.py file. Each plugin directory represents a plugin namespace.

plugins/
β”œβ”€β”€ botfather/
β”‚   β”œβ”€β”€ __init__.py
β”‚   └── cli.py
β”œβ”€β”€ devops/
β”‚   β”œβ”€β”€ __init__.py
β”‚   └── cli.py
└── your_plugin/
    β”œβ”€β”€ __init__.py
    └── cli.py

Transport

Create Your Own Skills

SMCP's plugin system makes it easy to give your AI new abilities. Each plugin is automatically discovered and its commands become new skills your AI can use.

Vibe Code Your Own Plugin

Plugins are simple Python scripts. Here's how easy it is to create one:

#!/usr/bin/env python3
"""
My Plugin CLI
"""
import argparse
import json
import sys

def main():
    parser = argparse.ArgumentParser(description="My Plugin CLI")
    subparsers = parser.add_subparsers(dest="command")
    
    cmd_parser = subparsers.add_parser("my-command", help="Execute my command")
    cmd_parser.add_argument("--param", required=True, help="Required parameter")
    
    args = parser.parse_args()
    
    if args.command == "my-command":
        result = execute_my_command(args.param)
        print(json.dumps(result))
    else:
        parser.print_help()
        sys.exit(1)

if __name__ == "__main__":
    main()

Available Plugins

πŸ€– BotFather

Telegram Bot API integration for sending messages and managing bots through SMCP.

βš™οΈ DevOps

Deployment and infrastructure management tools for system operations.

For comprehensive plugin development documentation, see the Plugin Development Guide.

Get Started in Minutes

Getting started with SMCP is simple. Install it, add a plugin, and your AI gains new skills instantly.

What You Need

Quick Start

# Clone the repository
git clone https://github.com/sanctumos/smcp.git
cd smcp

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Run the server
python smcp.py

The server will start on http://localhost:8000 by default with localhost-only access for security.

Configuration Options

Option Description
--host Host to bind to (default: 127.0.0.1 for security)
--port Port for the MCP server (default: 8000)
--allow-external Allow external connections (use with caution)

Built for Real Users, Trusted by Developers

SMCP was originally developed in the SanctumOS ecosystem as a Model Context Protocol server for the Letta AI framework, designed to make AI tools more capable and useful.

Through significant contributions from the Animus ecosystem, SMCP evolved into a robust, production-ready platform that makes it easy for anyone to enhance their AI's abilities.

While originally built for Letta, SMCP now works with the entire MCP ecosystem, giving users a simple way to connect their AI tools to the services and capabilities they need.

As an Animus Core Module, SMCP is the official, maintained implementation that bridges the gap between AI frameworks and the real worldβ€”making AI more useful for everyone.

Documentation & Resources

πŸš€ Getting Started

Complete setup guide and first steps with SMCP.

Read Guide β†’

🧩 Plugin Development

Learn how to build and deploy custom SMCP plugins.

Read Guide β†’

πŸ“‹ Examples

Practical examples and code samples for all use cases.

View Examples β†’

πŸ”§ API Reference

Complete API documentation and configuration options.

View API β†’

🚨 Troubleshooting

Common issues and solutions for all problems.

View Guide β†’

πŸ“¦ GitHub Repository

Source code, examples, and contributions.

Visit GitHub β†’