Visitar URL original
starting-ragchatbot-codebase/CLAUDE.md at main · jaychris/starting-ragchatbot-codebase · GitHub
Skip to content

CLAUDE.md

Latest commit

 

History

History
88 lines (65 loc) · 2.77 KB

CLAUDE.md

File metadata and controls

88 lines (65 loc) · 2.77 KB
Raw

CLAUDE.md

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

Project Overview

This is a RAG (Retrieval-Augmented Generation) chatbot for course materials. Users query course content through a web interface, and the system uses semantic search + Claude AI to generate answers.

Commands

# Install dependencies
uv sync

# Run the server (from project root)
cd backend && uv run uvicorn app:app --reload --port 8000

# Or use the convenience script
./run.sh

The app runs at http://localhost:8000 with API docs at http://localhost:8000/docs.

Architecture

Query Flow

  1. Frontend (frontend/script.js) → POST /api/query with {query, session_id}
  2. FastAPI (backend/app.py:56) → Routes to RAGSystem.query()
  3. RAGSystem (backend/rag_system.py:102) → Orchestrates the pipeline:
    • Gets conversation history from SessionManager
    • Calls AIGenerator.generate_response() with tools
  4. AIGenerator (backend/ai_generator.py:43) → Two-call pattern with Claude:
    • First call: Claude decides whether to use search_course_content tool
    • If tool used: executes search, then second call to synthesize answer
  5. CourseSearchTool (backend/search_tools.py:52) → Executes vector search
  6. VectorStore (backend/vector_store.py:61) → ChromaDB query with filters

Key Components

Component File Purpose
RAGSystem rag_system.py Main orchestrator, initializes all components
AIGenerator ai_generator.py Claude API calls, tool execution loop
VectorStore vector_store.py ChromaDB wrapper, two collections: course_catalog + course_content
DocumentProcessor document_processor.py Parses course files, sentence-based chunking
CourseSearchTool search_tools.py Tool definition for Claude, formats search results
SessionManager session_manager.py Per-session conversation history

Document Format

Course documents in docs/ follow this structure:

Course Title: [title]
Course Link: [url]
Course Instructor: [name]

Lesson 0: Introduction
Lesson Link: [url]
[content...]

Lesson 1: Topic
[content...]

Configuration

All settings in backend/config.py:

  • CHUNK_SIZE: 800 chars
  • CHUNK_OVERLAP: 100 chars
  • MAX_RESULTS: 5 search results
  • MAX_HISTORY: 2 conversation exchanges
  • EMBEDDING_MODEL: all-MiniLM-L6-v2
  • ANTHROPIC_MODEL: claude-sonnet-4-20250514

Data Models

Defined in backend/models.py:

  • Course → title (unique ID), lessons list, instructor, link
  • Lesson → number, title, link
  • CourseChunk → content with course/lesson metadata for vector storage

Environment

Requires .env file with:

ANTHROPIC_API_KEY=your_key_here