Semantic Workbench is a versatile tool for prototyping intelligent assistants and multi-agent systems, targeting developers and researchers. It offers a cohesive interface for creating, configuring, and interacting with one or multiple assistants, facilitating experimentation and measurement of agent behaviors.

How It Works

The system comprises a Python backend service, a React/TypeScript frontend UI, and pluggable assistant services. Assistants integrate via a RESTful API, allowing for language and framework agnostic development. This architecture promotes flexibility, enabling developers to use any language or platform for their assistants while leveraging a unified workbench for interaction and management.

Quick Start & Requirements

• Recommended: GitHub Codespaces for a turn-key cloud development environment.
• Local: Requires VS Code with Dev Containers or manual setup as per the setup guide.
• Running: Use VS Code's "Run and Debug" feature (semantic-workbench for the core, examples:python-01-echo-bot or examples:python-02-simple-chatbot for assistants).
• Access: Navigate to https://127.0.0.1:4000 in a web browser.
• Dependencies: Python, Node.js, Docker (for local dev containers). LLM API keys are optional for basic examples but required for advanced chatbots.

Highlighted Details

• Supports multi-assistant conversations within a single interface.
• Agent agnostic design allows integration of assistants built with any language or framework.
• Includes example assistants in Python and .NET, demonstrating echo bots, chatbots with moderation, and Azure AI integration.
• Provides tools for resetting service data and cleaning the development environment.

Maintenance & Community

Contributions are welcomed, governed by a Microsoft CLA. The project adheres to the Microsoft Open Source Code of Conduct.

Licensing & Compatibility

The project is licensed under the MIT License, permitting commercial use and integration with closed-source projects.

Limitations & Caveats

The project is primarily configured for VS Code and GitHub Codespaces, with local development requiring specific setup. While designed for flexibility, integrating custom assistant services requires adherence to defined protocols.