awesome-llm-apps: 100+ Open-Source AI Agent and RAG Apps

TL;DR: Shubhamsaboo/awesome-llm-apps is a 137,643-star curated index of more than a hundred open-source AI agent, agent skill, and RAG application examples in Python, all Apache 2.0 licensed and updated within the last 24 hours.

Shubhamsaboo/awesome-llm-apps is one of the largest curated indexes of working LLM applications on GitHub. The repo description promises “100+ AI Agents, Agent Skills and RAG Apps - Free and Open Source,” and the star count of 137,643 with 20,235 forks suggests the claim is accurate. The most recent push was September 12, 2026, so the index is not stale. The associated site theunwindai.com hosts longer-form articles that reference the same examples.

The repo is organized as a directory tree of standalone projects rather than a single monolithic codebase. Each example targets a specific use case: customer support agents, code review assistants, document Q&A pipelines, multi-agent orchestration patterns, retrieval-augmented chat over private corpora, and agent skills that bolt onto tools like Claude Code or Cursor. Topics on the repo include agents, llms, python, and rag, which gives you a sense of the breadth without diving into individual subdirectories.

For someone learning how to ship an LLM feature, the value is seeing ten different approaches to the same problem in one place. How does one team wire up retrieval over a PDF collection? How does another handle tool calling with structured outputs? How does a third design a multi-agent system where one agent plans and another executes? Reading the examples side by side reveals the trade-offs faster than any tutorial.

For production engineers, the Apache 2.0 license is the headline. You can lift any of the example code into a commercial project without legal review, modify it, and ship. The maintainer has done the work of keeping the dependency choices current, which means you are not inheriting a year-old example built on a deprecated SDK. Each subdirectory tends to have its own README, its own install instructions, and its own list of API keys required.

The collection is not exhaustive — there is no coverage of every framework or every agent pattern — but it is dense enough that you will find at least one reference implementation for most common tasks. If you are evaluating a new framework, the index is a useful place to start because the maintainers only include projects they have personally tested.

One caveat: the index grows quickly, and not every entry is production-quality. Some are weekend prototypes meant to demonstrate a pattern, not to ship to customers. The star and fork counts within each subdirectory are a better signal than the headline number on the root repo. Read the README of any example before assuming it is ready to deploy.

For Python developers specifically, the repo is a catalog of how to use modern LLM SDKs in practice. If you are new to OpenAI’s function calling, Anthropic’s tool use, or LangChain’s retrieval abstractions, you will find at least one minimal working example for each within the tree.

Source and Accuracy Notes

• Project page: https://github.com/Shubhamsaboo/awesome-llm-apps
• Source repository: Shubhamsaboo/awesome-llm-apps (verified via GitHub API html_url)
• License: Apache-2.0 (verified via GitHub API license.spdx_id)
• Stars at time of writing: 137,643