Adk MCP RAG
A Retrieval-Augmented Generation (RAG) system that leverages Google's Agent Development Kit (ADK) and Qdrant vector database via MCP server.
Installation
npx adk-mcp-ragAsk AI about Adk MCP RAG
Powered by Claude Β· Grounded in docs
I know everything about Adk MCP RAG. Ask me about installation, configuration, usage, or troubleshooting.
0/500
Reviews
Documentation
RAG Agent with Google ADK and Qdrant MCP server
A Retrieval-Augmented Generation (RAG) system that leverages Google's Agent Development Kit (ADK) and Qdrant vector database via MCP server.
Table of Contents
Overview
This project implements a Retrieval-Augmented Generation (RAG) system that combines the power of Google's Agent Development Kit (ADK) with Qdrant vector database (via MCP server) for efficient knowledge retrieval. The system enhances Large Language Model (LLM) responses by retrieving relevant context from a vector database before generating answers.
Architecture

Features
- Advanced Retrieval: Semantic search powered by Qdrant vector database
- Google ADK Integration: Leverages Google's Agent Development Kit for LLM capabilities
- MCP Server: Model Context Protocol server for Qdrant vector database
- Context Augmentation: Enhances LLM responses with relevant retrieved information
- Monitoring & Logging: Comprehensive tracking of system performance
Installation
Remember to install the uv first
# Clone the repository
git clone https://github.com/khoi03/adk-mcp-rag.git
cd adk-mcp-rag
# Create a virtual environment and install dependencies
uv sync
# Activate the virtual environment
# For macOS/Linux
source .venv/bin/activate
# For Windows
.venv\Scripts\activate
# Add required dependencies via requirement file (In case there are new libs)
uv add -r requirements.txt
Configuration
Environment Variables
Create a .env file in docker directory:
# Change directory into docker
cd docker
# Set up environment variables
cp .env.example .env
# Edit .env with your API keys and configuration
Example .env file:
# -------------------
# Google API keys
# -------------------
GOOGLE_API_KEY=YOUR_VALUE_HERE
# -------------------
# OPENAI API keys
# -------------------
OPENAI_API_KEY=YOUR_VALUE_HERE
# -------------------
# ANTHROPIC API keys
# -------------------
ANTHROPIC_API_KEY=YOUR_VALUE_HERE
# -------------------
# Network names
# -------------------
NETWORK_NAME=mcp-servers
# ----------------------------------
# Parameters for Qdrant MCP Server
# ----------------------------------
QDRANT_CONTAINER_NAME=qdrant-mcp
QDRANT_URL=http://qdrant:6333
QRANT_MCP_SSE=http://localhost:8888/sse
# QDRANT_LOCAL_PATH=/qdrant/db
COLLECTION_NAME=default_collection
QDRANT_EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
QDRANT_SEARCH_LIMIT=3
FASTMCP_HOST=0.0.0.0
FASTMCP_PORT=8888
Build Qdrant and Qdrant MCP
Build Qdrant and Qdrant MCP server using Docker Compose after completing the configuration steps:
# Build and start services
docker compose -p qdrant-mcp up --build -d
# Check running services
docker compose ps
# View logs
docker compose logs -f
# Stop and remove services
docker compose -p qdrant-mcp down
You can also access qdrant web UI at http://localhost:6333/dashboard
Usage
Document Ingestion
Push all your needed files into the data directory and run:
python local_vector_store/prepare_corpus_and_data_locally.py
Note: Currently only processes .md and .pdf files. The system will:
- Extract text from the documents
- Split the text into manageable chunks
- Generate embeddings for each chunk
- Store the embeddings in the Qdrant vector database
Basic Usage
To test and run the system with default settings:
python main.py
Built-in ADK-UI
For tracing, testing, and debugging with a UI, run the built-in web interface provided by ADK:
adk web
Project Structure
adk-mcp-rag/
βββ assets/ # Images and static files
βββ data/ # Documents for ingestion
βββ docker/ # Docker configurations
β βββ .env.example # Example environment variables
β βββ Dockerfile.qdrant # Docker file for qdrant mcp
β βββ docker-compose.yml # Docker Compose configuration
βββ agents/ # Main code
β βββ config/prompts.yml # Store prompts
β βββ tools/ # Embedding generation
β βββ mcp_tools.py # Manage MCP Tools
β βββ prompts.py # Manage Prompts
β βββ agent/ # Manage agents
βββ .gitignore # Git ignore file
βββ main.py # Main entry point
βββ README.md # This file
βββ requirements.txt # Python dependencies

