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

CLAUDE.md

Latest commit

 

History

History
89 lines (66 loc) · 2.72 KB

CLAUDE.md

File metadata and controls

89 lines (66 loc) · 2.72 KB
Raw

CLAUDE.md

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

Commands

Important: Always use uv to run the server and manage dependencies. Do not use pip directly. Make sure to use uv to manage all dependencies.

Install dependencies

uv sync

Run the application

cd backend
uv run uvicorn app:app --reload --port 8000

Or use the shell script: ./run.sh

Access points

Architecture

This is a RAG (Retrieval-Augmented Generation) chatbot that answers questions about course materials using Claude AI with tool-calling capabilities.

Request Flow

Frontend (vanilla JS) → POST /api/query → FastAPI (app.py)
    → RAGSystem.query() orchestrates:
        1. SessionManager: retrieves conversation history
        2. AIGenerator: calls Claude API with tools
        3. Claude decides whether to use search_course_content tool
        4. If tool used: CourseSearchTool → VectorStore → ChromaDB semantic search
        5. Claude synthesizes final response from search results
        6. SessionManager: saves exchange to history
    → Response with answer + sources returned to frontend

Backend Components (backend/)

File Role
rag_system.py Main orchestrator - coordinates all components
ai_generator.py Claude API client with tool execution loop
search_tools.py Tool definitions (CourseSearchTool) and ToolManager
vector_store.py ChromaDB wrapper with two collections: course_catalog (metadata) and course_content (chunks)
document_processor.py Parses course files, chunks text with sentence-aware splitting and overlap
session_manager.py Per-session conversation history
config.py Settings loaded from environment (chunk size, model names, etc.)

Document Format

Course files in docs/ must follow this structure:

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

Lesson 0: [Lesson Title]
Lesson Link: [url]
[Content...]

Lesson 1: [Lesson Title]
...

Key Configuration (backend/config.py)

  • CHUNK_SIZE: 800 characters per chunk
  • CHUNK_OVERLAP: 100 characters overlap between chunks
  • MAX_RESULTS: 5 search results per query
  • MAX_HISTORY: 2 conversation exchanges retained
  • ANTHROPIC_MODEL: claude-sonnet-4-20250514

Environment Variables

Required in .env file at project root:

ANTHROPIC_API_KEY=your_key_here

Git and Deployment

  • Never mention claude code contribution in anything (including but not limited to commits and PRs)