Building from scratch is hard. Use these templates to jumpstart your prototype.
- llm-basic — Simplest way to chat with an LLM using raw Python
requests. Start here if you're new!
- cv-traffic — Real-time vehicle detection logic using YOLOv8.
- data-processing — Cleaning and normalizing datasets for RL or Fine-Tuning.
- mcp-agent — Building an MCP-powered agent with tool-use capabilities. [Advanced/Optional]
- rag-pipeline — Dense retrieval pipeline using ChromaDB and Sentence-Transformers. [Advanced/Optional]
All kits require a .env file for API keys.
- Copy
.env.exampleto.env. - Add your keys (OpenAI, Anthropic, etc.).
- Run
pip install -r requirements.txt.