MCP Servers
The MCP (Model Context Protocol) Servers are integrations that extend the capabilities of AI assistants, allowing them to perform external actions, access APIs, interact with third-party services, and execute automations.
Overview
MCP Servers allow you to:
- Extend functionalities: Add capabilities to assistants through external tools
- Integrate services: Connect assistants to APIs and third-party services
- Automate tasks: Enable assistants to perform actions in external systems
- Access external data: Connect assistants to data sources and real-time services
What are MCP Servers?
MCP Servers are implementations of the Model Context Protocol that provide tools that can be used by AI assistants. Each MCP server offers a specific set of tools that expand the assistant's capabilities.
Examples of MCP Servers
- Brave Search: Allows assistants to perform web searches
- GitHub: Enables assistants to interact with Git repositories
- Playwright: Allows assistants to control web browsers
- Bible: Provides access to Bible verses
- YouTube Transcript: Extracts transcripts from YouTube videos
Accessing MCP Servers
To access MCP Servers:
- Navigate to the Projects section in the side menu
- Select the desired project
- Click on MCP Servers in the project’s side menu
Sources of MCP Servers
The Aumo platform allows you to install MCP servers from different sources:
Manual Installation
Add a JSON in the format defined by Anthropic with the necessary configurations to run a MCP STDIO, SSE, or streamable_http.
Docker Hub
Search for and install MCP servers available on Docker Hub. Hundreds of ready-to-use MCP servers are available.
GitHub
Search for and install MCP servers directly from a curated list on GitHub, allowing access to verified and functional servers.
Main Features
- Install servers: Add new MCP servers to the project
- Manage servers: View, configure, and manage installed servers
- Configure variables: Define the environment variables needed for each server
- View tools: Explore the available tools in each server
- Use in assistants: Select MCP servers when creating or editing assistants
Next Steps
- Install an MCP Server: Learn how to install MCP servers from Docker Hub or GitHub
- Manage MCP Servers: Find out how to manage installed servers
- Use MCP in Assistants: Learn how to integrate MCP servers into your assistants