QueryWeaver — Claude Code MCP server: setup
QueryWeaver turns plain-language database questions into SQL and returns results. Use it when an agent needs to inspect schemas and query data through MCP.
Open on GitHubWhat QueryWeaver does
QueryWeaver is an open-source MCP server for Text2SQL. It treats the database schema as a graph, accepts a natural-language question, and produces SQL with the result. Its MCP interface includes list_databases, connect_database, database_schema, and query_database.
Run the server in Docker on localhost:5000, then connect an MCP client to its HTTP surface at /mcp. The client configuration uses mcp.json, the server URL, and a Bearer token. QueryWeaver also exposes a REST API for schema graphs and streaming Text2SQL queries.
QueryWeaver connects to FalkorDB during startup, so it will not start without FalkorDB. Model access can use OPENAI_API_KEY, GEMINI_API_KEY, ANTHROPIC_API_KEY, or AZURE_API_KEY; configuration can be supplied through .env.
What it helps with
- turn a database question into SQL
- inspect tables, columns, and foreign keys
- run a Text2SQL query against a selected graph
- connect an MCP client to an HTTP server
- retrieve database schema data for an agent
Example prompts for the agent
- Show me the schema of the connected database and list its tables.
- Find the records that answer this question and run the SQL query.
- Connect to the database, select the right graph, and explain the query result.
How to install QueryWeaver
Install
The MCP registry has no launch command for it: see the repository README
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about QueryWeaver
How do I install QueryWeaver in Claude Code?
Start the server with docker run -p 5000:5000 -it falkordb/queryweaver, then add an HTTP MCP connection to http://127.0.0.1:5000/mcp in the client configuration. Provide a Bearer token for authorization.
What does QueryWeaver do?
It turns natural-language database questions into SQL using graph-based schema understanding, then returns the query and its result. The MCP operations cover database listing, connections, schema inspection, and querying.
What does QueryWeaver need to run?
QueryWeaver connects to FalkorDB at startup, so FalkorDB must be available before the server starts. Model access can use OPENAI_API_KEY, GEMINI_API_KEY, ANTHROPIC_API_KEY, or AZURE_API_KEY.
