Synechron Text2SQL MCP Server

Synechron Text2SQL MCP Server is an MCP server that provides natural language query interfaces for relational databases using large language models. It connects to multiple database engines, including PostgreSQL, MySQL, SQLite, and Microsoft Fabric, allowing developers, analysts, and enterprise teams to interrogate structured data without writing raw SQL. The server combines schema metadata and row-aware semantic retrieval to improve query precision and generate contextual SQL statements. It works with several LLM providers, including OpenAI, Azure OpenAI, AWS Bedrock, and local Ollama instances, while embedding schema information and sampled table rows to guide downstream generation. Beyond executing queries via tools like pandas or llama_index, the server synthesizes raw results into natural language summaries with optional Markdown formatting. Users can configure authentication using JSON Web Keys, integrate custom database definitions via JSON configuration files, and control row limits, retry thresholds, and sampling rates. By providing standardized Model Context Protocol tools, it enables clients such as Cursor, VS Code, and Claude Desktop to interact directly with internal databases across complex reporting and investigative workflows.

Category: Databases & Data Stores

Tags: database, natural-language, relational, SQL, text2sql

Visit Synechron Text2SQL MCP Server

How to install and configure Synechron Text2SQL MCP Server

1. Prepare your environment variables and configuration JSON (such as valves.json) defining your database connections, model selection, and indexing settings. 2. To register the server in an MCP client using standard I/O with Docker, add the following to your MCP configuration file (e.g., claude_desktop_config.json or Cursor settings): json { "mcpServers": { "text2sql": { "command": "docker", "args": [ "run", "-i", "--rm", "--env-file", ".env", "709825985650.dkr.ecr.us-east-1.amazonaws.com/synechron/mcp-text2sql:0.0.10-2025.07.04-rc3", "server" ] } } } 3. Alternatively, start the container as a streamable HTTP service with docker run -it -p8000:8000 ... server and configure your client to connect to http://localhost:8000/mcp. 4. Restart your MCP client to load the Text2SQL tools.

What you can do with Synechron Text2SQL MCP Server

• Querying relational databases like PostgreSQL, MySQL, or SQLite using plain English questions inside Cursor or Claude Desktop. - Generating automated summaries and Markdown tables from complex enterprise datasets such as cybersecurity incident logs or financial metrics. - Inspecting unfamiliar schemas and table structures using indexed vector embeddings of database metadata and sampled row values. - Executing analytical queries across Microsoft Fabric data stores with automated retries and configurable row result limits.

Key facts

• https://github.com/synechron-code/text2sql-mcp
• Databases & Data Stores, Developer Tools & Code Intelligence
• database, natural-language, relational, SQL, text2sql

Part of MCP Servers

What is Synechron Text2SQL MCP Server?

Synechron Text2SQL MCP Server is a Model Context Protocol server that translates natural language questions into SQL queries. It indexes database schemas and sample rows to help language models generate accurate queries across relational databases and return synthesized answers.

Which databases are supported by Synechron Text2SQL MCP Server?

The server supports PostgreSQL, MySQL, SQLite, and Microsoft Fabric databases. Users can define database connections, specify tables, set excluded system schemas, and supply custom prompt templates by passing connection strings through a dedicated JSON configuration file such as valves.json.

Which MCP clients work with Synechron Text2SQL MCP Server?

Any client compatible with the Model Context Protocol can connect to the server. Verified clients include Claude Desktop, Cursor, and VS Code, connecting either via stdio Docker execution or streamable HTTP at the designated server endpoint.

Which language models can I use with Synechron Text2SQL MCP Server?

The server works with multiple model providers, including OpenAI, Azure OpenAI, AWS Bedrock, and Ollama. You can configure both the primary text-to-SQL LLM and the embedding model used for schema and table indexing using environment variables or configuration JSON.