# MCP Server Getting Started Guide
Welcome to the Combined MCP Server. This guide will help you get started with both Redshift and Knowledgebase features.
## Installation
1. Clone the repository
2. Install dependencies with `uv pip install -e ".[dev]"`
3. Configure your environment variables in `.env.local`
4. Start the server with `python -m combined_mcp_server.main`
## Quick Start
### Redshift Tools
Use the `run_query` tool to execute SQL queries on your Redshift cluster:
```python
result = await run_query(
sql="SELECT * FROM users LIMIT 10",
db_user="analyst",
db_group="readonly"
)
```
### Knowledgebase Tools
Build your vector store from markdown files:
```python
result = await build_vectorstore()
```
Then query it with hybrid search:
```python
result = await query_vectorstore(
query="How do I configure the server?",
top_k=5,
search_type="hybrid"
)
```
## Configuration
All configuration is done through environment variables. See `.env.example` for available options.
### Required Variables
- `REDSHIFT_CLUSTER_ID` - Your Redshift cluster identifier
- `POSTGRES_SECRET_NAME` - Secrets Manager secret for PostgreSQL
- `KNOWLEDGEBASE_S3_BUCKET` - S3 bucket with markdown files
## Support
For issues or questions, please open an issue on GitHub.
