awesome-llm-apps

This repository is a curated library of 100+ open-source AI applications and templates, all Apache-2.0 licensed and described as hand-built and tested end-to-end. It is organized into practical categories so developers can quickly clone a working example, ship it, or sell it. Key sections include: - Agent Skills: reusable skills that can be installed into coding agents (Claude Code, Codex, Cursor) with a single npx command, such as Project Graveyard, Scope Creep Detector, Commit Archaeologist, and Dependency Doctor. - Starter AI Agents: single-file agents that run with just an API key, covering travel planning, data analysis, podcast generation, meme generation, music generation, medical imaging, and web scraping. - Advanced AI Agents: production-style agents with tools, memory, and multi-step reasoning, including deep research, fraud investigation, financial coaching, home renovation, product launch intelligence, and sales battle cards. - Always-on Agents: scheduled background agents that monitor sources and deliver briefings (Hacker News daily brief, release radar). - Multi-agent Teams: collaborating agent systems for finance, legal, recruitment, real estate, teaching, travel, game design, and UI/UX feedback. - Voice AI Agents: speech-in/speech-out agents for customer support, insurance claim intake, audio tours, and PDF voice Q&A. - Generative UI and Agentic Frontends: agents that render interactive components such as kanban boards, dashboards, and shadcn components. - Autonomous Game-Playing Agents: chess, tic-tac-toe, and 3D Pygame agents. - MCP AI Agents: integrations with browsing, GitHub, Notion, travel planning, and multi-MCP routing. - RAG Tutorials: a broad set of retrieval-augmented generation patterns, from basic chains to agentic, hybrid, multimodal, and knowledge-graph RAG, including local and cloud variants. - LLM Apps with Memory: persistent-memory chatbots for arXiv, travel, and general use. - Chat with X: conversational interfaces over GitHub, Gmail, PDFs, research papers, Substack, and YouTube. - LLM Optimization Tools: token and context-size reduction methods claiming 30-90% API cost savings. - LLM Fine-tuning: end-to-end recipes for Gemma 3 and Llama 3.2. - AI Agent Framework Crash Courses: deep dives into Google ADK and the OpenAI Agents SDK. Quick-start commands are provided for both installing agent skills into existing coding agents and running standalone agents locally, typically via Streamlit after a pip install. New templates are said to be added weekly, with tutorials published on the Unwind AI site. The repository is a well-organized reference for anyone building LLM-powered applications, from simple single-file demos to complex multi-agent orchestration.