OpenAgentPlatform/Dive

What it solves

Dive is a desktop application that allows users to connect various Large Language Models (LLMs) with external tools and data sources via the Model Context Protocol (MCP). It simplifies the process of adding capabilities to AI agents, removing the complexity of local server deployments and configuration.

How it works

Dive acts as an MCP Host, meaning it provides the interface and environment where LLMs (such as those from OpenAI, Anthropic, or Ollama) can execute functions. It supports both stdio and SSE modes for MCP integration. Users can either install MCP servers locally or use OAP Cloud via OAPHub.ai for one-click access to managed servers without needing local dependencies like Python or Docker.

Who it’s for

It is designed for users who want a cross-platform desktop AI agent interface that can interact with their local files, execute bash commands, and integrate with a wide variety of third-party tools through the MCP standard.

Highlights

• Universal LLM Support: Works with ChatGPT, Anthropic, Ollama, and OpenAI-compatible models.
• MCP Integration: Seamlessly connects to MCP servers via stdio or SSE, including an intelligent agent to help with installation.
• Built-in Local Tools: Comes pre-configured with a File Manager, Bash execution, and Fetch for web requests.
• Cloud-Managed Tools: Integration with OAPHub.ai for zero-configuration access to managed MCP servers.
• Cross-Platform: Available for Windows, macOS, and Linux with both Tauri and Electron architectures.
• Customization: Supports personalized system prompts, granular tool control, and multi-language support for 24+ languages.