RAGify Docs
A Developers Tool — Scrape entire documentation recursively and ask questions using AI
RAGify Docs API 📚🤖
A comprehensive API for scraping documentation from any URL and asking intelligent questions about it using Retrieval-Augmented Generation (RAG) powered by LangChain and Groq.
🎯 Overview
RAGify Docs API is a developer-friendly tool that automates the process of extracting knowledge from documentation websites. Instead of manually reading through documentation, you can now:
- 🔗 Provide a documentation URL
- 🤔 Ask any question about the content
- 🧠 Get intelligent, sourced answers powered by AI
Built with LangChain, FastAPI, and Groq's powerful LLM, this tool handles the heavy lifting of document loading, chunking, embedding, retrieval, and generation—all with caching for performance.
🚀 Quick Start
Prerequisites
- Python 3.12 or higher
- Groq API key (free tier available at console.groq.com)
Installation
-
Clone or navigate to the project directory:
cd RAGify-Docs-API -
Create a virtual environment:
python -m venv .venv .venv\Scripts\activate # On Windows # or source .venv/bin/activate # On macOS/Linux -
Install dependencies:
pip install -r requirements.txt # or pip install -e . -
Set up environment variables: Create a
.envfile in the project root:GROQ_API_KEY=your_groq_api_key_here
📖 Usage
Option 1: FastAPI Server
Start the API server:
uvicorn app:app --reload
The API will be available at http://localhost:8000
API Documentation:
- Interactive Docs:
http://localhost:8000/docs(Swagger UI) - Alternative Docs:
http://localhost:8000/redoc
Option 2: Interactive CLI
Run the interactive mode:
python main.py
You'll be prompted to:
- Enter a documentation URL (e.g.,
https://docs.langchain.com/oss/python/langchain/overview) - Ask questions about the scraped documentation
- View answers with source URLs
🔌 API Endpoints
GET /
Welcome endpoint with basic information.
Response:
{
"message": "Welcome to the RAGify Docs API. Use the /ragify endpoint to ask questions about documentation from a given URL."
}
POST /ragify
Main endpoint for asking questions about documentation.
Request Body:
{
"url": "https://docs.langchain.com/oss/python/langchain/overview",
"query": "What is LangChain?"
}
Response:
{
"answer": "LangChain is a framework for developing applications powered by language models...",
"sources": [
"https://docs.langchain.com/oss/python/langchain/overview",
"https://docs.langchain.com/oss/python/langchain/guides"
]
}
Status Codes:
200— Success500— Error (RAG initialization or chain invocation failed)
🏗️ Architecture
┌─────────────────────────────────────────────┐
│ FastAPI Server │
│ (app.py with CORS) │
└────────────┬────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────┐
│ URL Caching Layer │
│ (Avoid re-processing same URLs) │
└────────────┬────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────┐
│ RAG Pipeline (main.py) │
├─────────────────────────────────────────────┤
│ 1. RecursiveUrlLoader → Scrape docs │
│ 2. RecursiveCharacterTextSplitter │
│ 3. HuggingFace Embeddings │
│ 4. InMemoryVectorStore │
│ 5. MMR Retriever (k=5, diverse results) │
│ 6. ChatGroq LLM (openai/gpt-oss-120b) │
│ 7. LangChain RAG Chain │
└─────────────────────────────────────────────┘
Component Details
| Component | Purpose |
|---|---|
| RecursiveUrlLoader | Crawls documentation pages recursively with custom HTML extraction |
| RecursiveCharacterTextSplitter | Splits documents (1000 chars/chunk, 200 char overlap) |
| HuggingFace Embeddings | Converts text to vectors using sentence-transformers/all-MiniLM-L6-v2 |
| InMemoryVectorStore | Stores embeddings for semantic search |
| MMR Retriever | Returns diverse, relevant documents (5 best from 10 candidates) |
| ChatGroq | LLM for generating answers (model: openai/gpt-oss-120b, temp: 0.2) |
💻 Example Usage
Using cURL
curl -X POST "http://localhost:8000/ragify" \
-H "Content-Type: application/json" \
-d '{
"url": "https://docs.langchain.com/oss/python/langchain/overview",
"query": "What is the purpose of LangChain?"
}'
Using Python Requests
import requests
response = requests.post(
"http://localhost:8000/ragify",
json={
"url": "https://docs.langchain.com/oss/python/langchain/overview",
"query": "How do I create a RAG chain?"
}
)
print(response.json())
Using JavaScript/Fetch
fetch("http://localhost:8000/ragify", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({
url: "https://docs.langchain.com/oss/python/langchain/overview",
query: "What are chains in LangChain?"
})
})
.then(res => res.json())
.then(data => console.log(data));
⚙️ Configuration
Environment Variables
Create a .env file to configure the application:
# Groq API Key (required)
GROQ_API_KEY=your_api_key_here
# Optional: Customize embedding model
# EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
# Optional: Customize LLM model
# LLM_MODEL=openai/gpt-oss-120b
Tunable Parameters (in main.py)
# Chunk size and overlap
text_splitter = RecursiveCharacterTextSplitter(
chunk_size=1000, # Increase for longer context
chunk_overlap=200 # Increase for better continuity
)
# Retriever settings
retriever = vector_store.as_retriever(
search_type="mmr",
search_kwargs={
"k": 5, # Number of results to return
"fetch_k": 10, # Candidates to consider
"lambda_mult": 0.5 # 1.0 = relevance, 0.0 = diversity
}
)
# LLM settings
llm = ChatGroq(
model="openai/gpt-oss-120b",
temperature=0.2 # Lower = factual, Higher = creative
)
🛠️ Development & Debugging
Run with hot-reload
uvicorn app:app --reload
Run tests (if added)
pytest
Check for issues
# Verify dependencies
pip check
# Lint code
pylint app.py main.py
🎉 Happy RAGifying
Build smarter, faster, and more informed applications with RAGify Docs API.
RAGify Docs API — Because great documentation deserves great answers.
相关服务器
Bright Data
赞助Discover, extract, and interact with the web - one interface powering automated access across the public internet.
Apify
Extract data from any website with thousands of scrapers, crawlers, and automations
Sports Trading Card Agent
Real-time sports card pricing, market analysis, arbitrage detection, grading ROI, investment advice, and player stats (NBA/NFL/MLB). 9 tools for AI agents helping collectors and investors.
Web Scout
An MCP server for web search and content extraction using DuckDuckGo.
YouTube Transcript Extractor
Extracts transcripts from public YouTube videos.
Web Fetch
Fetches and transforms web content, including JavaScript-rendered pages and media files, into various formats.
MCP Server Collector
Discovers and collects MCP servers from the internet.
Plasmate MCP
Agent-native headless browser that converts web pages to structured Semantic Object Model (SOM) JSON -- 4x fewer tokens than raw HTML with lower latency on Claude and GPT-4o.
brosh
A browser screenshot tool to capture scrolling screenshots of webpages using Playwright, with support for intelligent section identification and multiple output formats.
Playwright
Provides browser automation capabilities using Playwright. Interact with web pages, take screenshots, and execute JavaScript in a real browser environment.
Read URL MCP
Extracts web content from a URL and converts it to clean Markdown format.