Most curated lists show links; this one stitches concrete, runnable app examples into a learning path for building agentic systems and retrieval-augmented apps. Instead of only naming tools, the repo groups end-to-end examples (starter agents, RAG, voice, MCP, memory) with course playlists and reproducible starters so you can go from concept to prototype without hunting for missing pieces.
What Sets It Apart
- Breadth with runnable examples: over 80 projects spanning starter agents, simple/advanced agent patterns, RAG demos, voice pipelines, MCP examples, and memory-enabled assistants—each entry links to a concrete repository or demo so you can clone and run a full flow. This means less discovery time and more hands-on experimentation.
- Learning + reference combo: playlists and course links are curated alongside code, so the repo functions both as a tutorial index (step-by-step videos) and a reference library of implementation patterns (LangChain/LlamaIndex, MCP, Memori, Nebius integrations).
- Focus on agent workflows and production patterns: many examples emphasize orchestration, multi-agent patterns, observability and security (MCP servers, Docker sandboxing, telemetry), making it useful for developers thinking beyond single-turn chatbots.
Who it's for & tradeoffs
Great fit if you are a developer or engineering team wanting reproducible examples to learn agent design, RAG architectures, or voice/real-time pipelines—especially when you prefer runnable starters and video guidance. Look elsewhere if you need a single opinionated framework or polished SaaS; this repo is a broad cookbook rather than a turnkey product and requires reading each project's README and wiring API keys for many demos.
Where It Fits
This is a developer-facing curated cookbook: not a research paper or commercial product, but a practical bridge between tutorials and production patterns. Use it to prototype ideas, compare approaches across frameworks, or assemble building blocks for production systems.
Quick usage note
The repository assumes familiarity with Python tooling and common AI providers (API keys required for many demos). Many entries are provider-agnostic but expect developers to supply credentials and follow each subproject's README to run examples.
