Awesome AI Agents › Educational & Learning Agents
Inno Agent is a single-learner study companion built on top of the Pi coding-agent SDK without modifying its kernel, adding all learning behaviour through registered tools and one extension hook so the runtime stays compatible with upstream. Its central design idea is layered memory instead of a flat chat summary. L1 is a learner profile holding goals, knowledge states, misconceptions and preferences, summarised into a context pack injected on each turn and open for the learner to inspect and correct. L2 is a native wiki knowledge base of human-readable, agent-queryable pages with hybrid retrieval combining BM25 lexical search and a knowledge graph, LLM-assisted summarisation, and ingestion of PDF, Office and image files. L3 indexes session history in SQLite with FTS5 and performs threshold-gated retrieval across past conversations. Around the memory layers sits a learning loop: a cron scheduler whose jobs can be created in natural language and run from the agent, the UI or a daemon, personal instant-messaging channels for Feishu and WeChat with a unified reminder dispatcher, and a Practice Lab providing a workspace-scoped browser terminal over xterm.js and WebSocket whose run records the agent can read. The project ships in three forms sharing the same runtime state: an Electron desktop application for macOS and Windows, a web UI built with React 19 and Tailwind 4 served by a Node HTTP server with SSE streaming, and a terminal CLI. A simple mode with preset workspaces targets non-technical users, and a skill system imports skills and presets from a remote hub. Any OpenAI-completions or Anthropic-messages endpoint can be used as the model provider, switchable live. The design is deliberately single-user: one process runs one active session, there is no authentication, tenancy or horizontal scaling, and concurrent cross-session calls return a busy status rather than queueing. TypeScript, licensed MIT.
https://github.com/hhyqhh/inno-agent
AI Agents for Beginners is a Microsoft educational course with 10 lessons teaching the fundamentals and practical skills to build AI agents using Azure AI and GitHub models, supported by multi-language resources and community engagement.
This repository offers a comprehensive, hands-on curriculum (503 lessons across 20 phases) for building AI systems from scratch, covering math to autonomous agents, using Python, TypeScript, Rust, ...
A comprehensive collection of free resources to learn AI, machine learning, large language models, and AI agents from scratch.
A comprehensive educational repository providing clear, step-by-step Python implementations and explanations of a wide range of reinforcement learning algorithms for interactive learning and deep understanding.
TheoremExplainAgent is an AI system that generates long-form Manim videos to visually explain mathematical theorems, enhancing understanding through multimodal explanations combining text, visuals, and audio.
LangChain-OpenTutorial is an open, community-driven tutorial resource providing comprehensive learning materials and practical examples for LangChain and LangGraph technologies, suitable for users of all skill levels worldwide.
Groundhog is an AI coding assistant designed as a teaching tool to help users understand and leverage the inner workings of coding agents like Cursor through detailed code explanations and a modern Rust-based CLI.
An educational repository with code and notebooks accompanying the "Quick Start Guide to Large Language Models - Second Edition" book, covering practical applications and advanced techniques for large language models.