Aci

ACI.dev is an open-source AI agent tool infrastructure that provides multi-tenant authentication, permission management, and dynamic discovery functions for over 600 pre-integrated tools. It supports calls through a unified MCP Server or a lightweight SDK, simplifying the development of production-level AI agents.

Developer tools Artificial intelligence chatbots #AI Agent #Tool Integration #Multi-tenant Authentication #Unified Interface .Python

rating : 4 points

downloads : 6.7K

update time : 2025-04-28

Open Site

What is the ACI.dev MCP Server?

ACI.dev's Unified MCP Server is an intermediate layer that enables AI agents to securely access over 600 tools and services through a unified interface. It handles complex authentication, permission management, and tool discovery, allowing developers to focus on building agent logic.

How to use the MCP Server?

You can directly call tool functions through our Python SDK or integrate with your LLM framework via the Unified MCP Server. The server provides REST API and WebSocket interfaces.

Use Cases

Suitable for AI agent scenarios that require tool usage, such as personal assistants, research assistants, sales automation, customer support, etc. Particularly suitable for production environments that require multi-user and multi-service authentication.

Main Features

Integration of 600+ Tools

Pre-integrated with popular services such as Google Calendar, Slack, and Notion, eliminating the need to develop API clients separately.

Unified Interface

Provides a standardized access method through the MCP Server or SDK, simplifying agent development.

Multi-tenant Authentication

Built-in OAuth process and key management, supporting authentication for developers and end-users.

Dynamic Tool Discovery

Agents can dynamically discover available tools, avoiding context window overload.

Advantages

Eliminates the infrastructure burden of developing each tool integration

Provides a production-ready authentication and permission system

Framework-independent, can be used with any LLM

100% open-source, no vendor lock-in

Limitations

Requires deploying and maintaining your own server instance

Some advanced tools may require additional configuration

Integrating new tools may require community contributions

How to Use

Install the Python SDK

Install our SDK package via pip

Initialize the Client

Create an SDK client instance and configure authentication

Call Tools

Call any available tool through the unified interface

Usage Examples

Personal Assistant Schedules a Meeting

The agent creates a meeting on Google Calendar based on the user's request

Research Assistant Saves Results

The research agent saves the found information to a Notion database

Frequently Asked Questions

What's the difference between the MCP Server and direct API calls?

How to add custom tools?

Does it support enterprise-level SLA?

Related Resources

Official Documentation

Complete product documentation and API reference

Python SDK Repository

SDK source code and usage examples

Demo Video

Function demonstration of the Unified MCP Server

Discord Community

Join our developer community

🚀 ACI: Open-Source Infrastructure for Unified MCP Servers

ACI (Artificial Intelligence Infrastructure) is an open-source platform that provides infrastructure support for building and managing intelligent agents. It integrates various functional modules and services through a unified MCP (Multi-Capability Processor) server architecture, enabling developers to easily build intelligent systems capable of handling complex tasks.

🚀 Quick Start

⚠️ Important Note

This repository is for the ACI.dev platform. If you're looking for the Unified MCP servers built with ACI.dev, please check out aci-mcp.

✨ Features

Unified Interface: Offers a unified API interface to simplify integration with various services and data sources.
Scalability: Supports a plug-in architecture, facilitating the addition of new functional modules and services.
High Availability: Adopts a distributed design to ensure system stability and reliability.
Multi-Language Support: Supports multiple programming languages, allowing developers to choose their familiar languages for development.

📚 Documentation

Project Structure

.
├── backend          # Backend service implementation, including API interfaces and business logic
│   ├── README.md     # Backend operation guide
│   └── ...           # Other backend-related files
├── frontend         # Frontend interface for developers to manage and configure services
│   ├── README.md     # Frontend operation guide
│   └── ...           # Other frontend-related files
└── integrations      # Integrated third-party services and functional modules
    ├── examples       # Example integration implementations
    └── ...            # Other integration-related files

🤝 How to Contribute

We welcome community members to actively participate in the development of ACI! Please refer to CONTRIBUTING.md to learn how to contribute to the project.

❓ Issues and Support

If you encounter any problems during use or need assistance, please contact us through the following channels:

GitHub Issues: Submit issues to the Issues page.
Discord Community: Join our Discord server to communicate with other developers.
Mailing List: Subscribe to our mailing list to get project updates and discussions.

📄 License

This project is licensed under the Apache 2.0 license. Please refer to the LICENSE file for details.

ACI is an open and collaborative project aiming to promote the development of artificial intelligence infrastructure. Welcome every interested developer to join us and build a powerful and flexible intelligent agent platform together!