AgentVerse is an open-source platform designed for multi-AI agent collaborative dialogue, enabling teams of AI experts with diverse specializations and personalities to autonomously interact and solve problems. It targets users seeking multi-angled professional insights, creative assistance, decision support, learning resources, or code development aid, offering a benefit of leveraging a simulated expert team for comprehensive solutions.

How It Works

The platform facilitates collaboration through autonomous dialogue management, where AI agents, each with defined roles, personalities, and expertise, communicate and discuss tasks without manual intervention. It integrates a powerful tool system supporting the Model Context Protocol (MCP), allowing agents to interact with external functionalities such as file operations, data analysis, information retrieval, and code execution. An embedded file management system, built on LightningFS, further enhances agent capabilities by enabling direct file manipulation. This approach simulates a human expert team, automating coordination and extending AI functionality beyond basic chat interactions.

Quick Start & Requirements

• Environment Requirements: Node.js >= 18, pnpm >= 8.0.
• Installation:
  1. git clone https://github.com/Peiiii/AgentVerse.git
  2. cd AgentVerse
  3. pnpm install
  4. Configure environment variables by copying .env.example to .env and setting AI service provider and API keys.
  5. pnpm dev
• Access: The development server is accessible at http://localhost:3000.
• Links: GitHub Repository

Highlighted Details

• Pre-set Expert Teams: Features 14 pre-configured teams for diverse scenarios including thinking, creative writing, business analysis, and product development.
• Custom Agent Creation: Supports creating bespoke AI agents through conversational interfaces or an intelligent configuration assistant, allowing flexible customization of roles, personalities, and expertise.
• MCP Tool Integration: Enables agents to leverage external tools for file system operations, data querying, information retrieval, task automation, and code generation/execution.
• File Management System: An integrated system based on LightningFS provides capabilities for file/directory CRUD operations, content editing, previewing, searching, and upload/download.
• Responsive Design: Offers a seamless user experience across both desktop and mobile devices, with support for dark and light themes.
• Extensible Architecture: Designed with extensibility in mind, supporting plugins and custom extensions.

Maintenance & Community

The primary channel for community interaction and issue reporting is GitHub Issues.

Licensing & Compatibility

The project is licensed under the permissive MIT License, which generally allows for commercial use and integration into closed-source projects without significant restrictions.

Limitations & Caveats

The provided README does not explicitly detail known limitations, unsupported platforms, or alpha/beta status. Features listed as completed in the development plan suggest a stable implementation for those functionalities.