A complete starter kit for building custom AI agents using the Model Context Protocol (MCP). Includes a minimal MCP server, OpenAI integration, and sample tools to help developers understand and extend AI-driven workflows. Short tagline (for repo header): 🚀 Build and run your own MCP-powered AI agents — fast, modular, and open-source.

An extension for the alfredapp/openai-workflow that enhances functionality by simplifying prompt customization for various use cases. It requires the original workflow and offers an easy way to tailor ChatGPT for specific tasks or creative needs, giving users more flexibility and control in Alfred.

Extending OpenAI’s Whisper with performance optimizations and seamless integration into LLM-driven workflows for more accurate, contextual, and intelligent speech-to-text applications.

An agentic workflow to extend Duolingo streak (reimplemented with OpenAI Agents SDK, leading to better performance at https://github.com/Brofer-The-First/duolingo-streak-agents-sdk)

A multi-agent AI personal assistant built with n8n and OpenAI, integrated with Telegram. It automates tasks like scheduling, expense tracking, email, contact updates, and more using natural language. Easily extendable for more agents and workflows.

An AI-powered Telegram assistant built with n8n, integrating OpenAI, SerpAPI, Google Calendar, and Google Tasks. It enables natural language chat, task management, event scheduling, and web search all from Telegram AI Agent chat. The modular workflow is easy to customize and extend.