Atlantis FAQs

Question 1

What is Atlantis?

Accepted Answer

Atlantis is a Python-based host (often called a 'remote' or 'server') for the Model Context Protocol (MCP) ecosystem. It enables you to run dynamic Python functions and integrate third-party MCP tools directly within your local environment.

Question 2

What is the Model Context Protocol (MCP)?

Accepted Answer

MCP is a protocol designed to invert traditional cloud architectures, promoting more local processing and control. It gives bots and AI agents enhanced capabilities by facilitating local tool execution and context management.

Question 3

What types of tools can Atlantis host?

Accepted Answer

Atlantis can host two main types of tools: 'Dynamic Functions' (custom Python functions that act as tools, with automatic reloading on modification) and 'Dynamic MCP Servers' (third-party MCP tools installed via JSON configuration files).

Question 4

Can Atlantis be used standalone or with a cloud service?

Accepted Answer

Yes, Atlantis offers flexibility. It can operate as a standalone MCP server locally, or you can connect it to the experimental Atlantis cloud service for additional features like managing multiple instances and sharing tools.

Question 5

How does Atlantis integrate with existing MCP clients?

Accepted Answer

Atlantis simplifies integration with popular MCP clients such as Windsurf, Cursor, or Claude Code. You can configure these clients to connect to your Atlantis remote, allowing them to leverage the dynamic functions and third-party tools you've installed.

Atlantis

Atlantis

Key Features

Use Cases

Key Features

Use Cases