Skip to content

CLAUDE.md

Latest commit

 

History

History
54 lines (40 loc) · 1.59 KB

CLAUDE.md

File metadata and controls

54 lines (40 loc) · 1.59 KB
Raw

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Repository Overview

Educational course repository for "Systematically Improving RAG Applications" - teaches data-driven approaches to building and improving Retrieval-Augmented Generation (RAG) systems.

Common Commands

# Package management (always use uv)
uv install                    # Install dependencies
uv add <package>              # Add new package
uv sync                       # Sync dependencies

# Code quality
uv run ruff check --fix --unsafe-fixes .
uv run ruff format .

# Documentation
mkdocs serve                  # Local dev server at localhost:8000
mkdocs build                  # Build static site

# Tests
uv run pytest -q              # Run tests

Project Structure

latest/           # Current course version (weeks 0-6, case study)
cohort_1/         # Previous cohort materials (reference)
cohort_2/         # Previous cohort materials (reference)
docs/             # MkDocs documentation
  workshops/      # Main ebook content (chapters 0-7)
  office-hours/   # Q&A session summaries
  talks/          # Industry expert presentations

Code Style

  • Python 3.11, type hints throughout
  • Async-first for I/O operations
  • typer + rich for CLI tools
  • pydantic for validation
  • 9th-grade reading level for educational content
  • No emojis in code or docs

Key Patterns

  • RAG Flywheel: Measure → Analyze → Improve → Iterate
  • Evaluation-first: Always measure before and after changes
  • Synthetic data: For cold-start problems