Quick Start Guide

# Create workspace directory
mkdir my-qdrant-workspace
cd my-qdrant-workspace

# Create .env file with your credentials
cat > .env << EOF
QDRANT_URL=http://localhost:6333
OPENAI_API_KEY=your-openai-api-key-here
QDRANT_COLLECTION_NAME=quickstart
EOF

# Initialize workspace with default configuration
qdrant-loader --workspace . init

# Expected output:
# ✅ Collection initialized successfully: quickstart

# Check project status
qdrant-loader project --workspace . status

# Expected output shows project configuration and connection status

# Create a sample document
cat > sample-doc.md << EOF
# Welcome to QDrant Loader

QDrant Loader is a powerful tool for ingesting documents into vector databases.

## Key Features
- Multi-source data ingestion
- 20+ file format support
- AI tool integration via MCP
- Intelligent text chunking

## Use Cases
- Knowledge base creation
- Document search and retrieval
- AI-powered development workflows
EOF

# Create a basic configuration file
cat > config.yaml << EOF
projects:
  quickstart:
    display_name: "Quick Start Project"
    description: "Getting started with QDrant Loader"
    collection_name: "quickstart"
    sources:
      localfile:
        sample_docs:
          path: "."
          include_patterns: ["*.md"]
          recursive: false
EOF

# Ingest the document
qdrant-loader --workspace . ingest

# Expected output:
# 📄 Processing documents from configured sources
# ✅ Ingested: 1 document, 4 chunks
# 🔍 Collection: quickstart

# Update config.yaml to include git source
cat > config.yaml << EOF
projects:
  quickstart:
    display_name: "Quick Start Project"
    description: "Getting started with QDrant Loader"
    collection_name: "quickstart"
    sources:
      git:
        qdrant_docs:
          url: "https://github.com/qdrant/qdrant-client"
          include_patterns: ["*.md", "*.rst"]
          exclude_patterns: ["node_modules/", ".git/"]
EOF

# Ingest the repository
qdrant-loader --workspace . ingest

# Expected output:
# 📁 Cloning repository...
# 📄 Processing: multiple files found
# ✅ Ingested: multiple documents and chunks
# 🔍 Collection: quickstart

# Create a sample project structure
mkdir -p my-project/docs
cat > my-project/docs/overview.md << EOF
# My Project Overview
This is a sample project for testing QDrant Loader.
EOF

cat > my-project/docs/api.md << EOF
# API Documentation
Our API provides powerful search capabilities.
EOF

# Update config.yaml to include the directory
cat > config.yaml << EOF
projects:
  quickstart:
    display_name: "Quick Start Project"
    description: "Getting started with QDrant Loader"
    collection_name: "quickstart"
    sources:
      localfile:
        project_docs:
          path: "my-project/"
          include_patterns: ["*.md"]
          recursive: true
EOF

# Ingest the entire directory
qdrant-loader --workspace . ingest

# Expected output:
# 📁 Scanning directory: my-project/
# 📄 Processing: 2 files found
# ✅ Ingested: 2 documents, multiple chunks
# 🔍 Collection: quickstart

# Check project status
qdrant-loader project --workspace . status

# List configured projects
qdrant-loader project --workspace . list

# Start MCP server (keep this terminal open)
mcp-qdrant-loader

# Expected output:
# 🚀 QDrant Loader MCP Server starting...
# 📡 Server running on stdio
# 🔍 Available tools: search, hierarchy_search, attachment_search
# ✅ Ready for connections

{
  "name": "qdrant-loader",
  "command": "mcp-qdrant-loader",
  "args": [],
  "env": {
    "QDRANT_URL": "http://localhost:6333",
    "OPENAI_API_KEY": "your-openai-api-key-here",
    "QDRANT_COLLECTION_NAME": "quickstart"
  }
}

Can you search for information about QDrant Loader features?

1. "What are the key features mentioned in the documentation?"
2. "Find information about API endpoints"
3. "Search for installation instructions"
4. "What file formats are supported?"

# Query the database directly (if available)
python packages/qdrant-loader/tests/scripts/query_qdrant.py \
  --config config.yaml \
  --env .env \
  --search "QDrant Loader features"

# Check if QDrant is running
curl http://localhost:6333/health

# Start QDrant with Docker if not running
docker run -p 6333:6333 -p 6334:6334 qdrant/qdrant

# Verify connection
qdrant-loader project --workspace . status

# Check API key
echo $OPENAI_API_KEY

# Test API key directly
curl -H "Authorization: Bearer $OPENAI_API_KEY" \
  https://api.openai.com/v1/models

# Update .env file with correct key

# Check file path exists
ls -la sample-doc.md

# Check configuration file
cat config.yaml

# Use verbose mode for debugging
qdrant-loader --workspace . --log-level DEBUG ingest

# Verify MCP server is running
ps aux | grep mcp-qdrant-loader

# Check MCP server logs
mcp-qdrant-loader --log-level DEBUG

# Restart MCP server
pkill -f mcp-qdrant-loader
mcp-qdrant-loader

# Verify documents are ingested
qdrant-loader project --workspace . status

# Check collection status
qdrant-loader project --workspace . list

# Re-ingest if needed
qdrant-loader --workspace . ingest --log-level DEBUG