Universal MCP-Server that automatically generates REST & MCP APIs for your SQL/warehouse data – optimized for LLMs and AI agents.
https://github.com/centralmind/gatewayYour AI agents need database access, but building secure, well-documented APIs takes weeks. CentralMind Gateway analyzes your database schema and automatically generates production-ready REST and MCP APIs in minutes – no manual endpoint definition required.
You're building AI agents that need to query your PostgreSQL analytics database or pull customer data from MySQL. The traditional approach means weeks of API development: defining endpoints, writing SQL queries, handling parameters, securing access, and documenting everything.
Meanwhile, your AI agent is waiting, and your sprint deadline isn't moving.
Gateway connects to your database, analyzes the schema and sample data, then uses an LLM to generate optimized API configurations automatically:
docker run --platform linux/amd64 -p 9090:9090 \
ghcr.io/centralmind/gateway:v0.2.18 start \
--connection-string "postgres://user:pass@host:5432/analytics"
Within seconds, you get:
http://localhost:9090/http://localhost:9090/sseBefore Gateway:
With Gateway:
Your Claude Desktop, Cursor, or custom LangChain agents get instant database access through MCP. No more hardcoded database connections in your agent code.
{
"mcpServers": {
"database": {
"command": "docker",
"args": ["run", "-i", "--platform", "linux/amd64",
"ghcr.io/centralmind/gateway:v0.2.18",
"--connection-string", "postgres://...", "start", "stdio"]
}
}
}
Connect your data warehouse (Snowflake, BigQuery, ClickHouse) and get queryable APIs for building custom analytics tools. Gateway handles the complexity of different SQL dialects and optimizations.
Need to expose your database for a hackathon project or proof-of-concept? Gateway gets you from database schema to working API faster than you can write the first endpoint manually.
Gateway isn't just fast – it's secure by default:
Works with your existing infrastructure:
Gateway implements MCP properly, so it integrates seamlessly with:
export GEMINI_API_KEY='your-key' # Free tier available
./gateway discover \
--ai-provider gemini \
--connection-string "postgresql://user:pass@host/db" \
--prompt "Generate readonly API for customer analytics"
Review the generated gateway.yaml with your optimized endpoints
Start the server:
./gateway start --config gateway.yaml
Your API is live with documentation, security, and AI agent compatibility built-in.
Building database APIs manually means reinventing solutions for parameter validation, SQL injection prevention, response caching, and documentation generation. Gateway delivers all of this automatically, optimized specifically for AI agent workloads.
The 405 GitHub stars and active development show this isn't experimental – it's production-ready infrastructure that other developers are already using to ship faster.
Stop writing boilerplate database APIs. Generate them instead.