QuerySculptor
@flinstonedev
QuerySculptor is a GraphQL query builder designed for AI agents that provides 26 structured tools via the Model Context Protocol (MCP). Instead of generating complex GraphQL queries from scratch, AI agents can build queries step-by-step with real-time validation, schema introspec
Overview
What is QuerySculptor?
QuerySculptor is a server application that helps AI agents interact with GraphQL APIs through structured tools instead of raw query generation. It provides 26 tools via the Model Context Protocol (MCP) for building, validating, and executing GraphQL queries step-by-step. It is designed for developers using AI agents like Cursor.
How to use QuerySculptor?
Add the public demo endpoint to your Cursor mcp.json configuration. For local setup, clone the repository, run pnpm install, copy example.env to .env and customize, then launch with vercel dev. The server becomes available at http://localhost:3000/mcp.
Key features of QuerySculptor
- 26 MCP tools for building GraphQL queries step-by-step
- Step-by-step query building instead of raw string generation
- Pre-configured public demo using the Pokemon API
- Supports complex queries like union types and nested fields
- Easily deployable locally via Vercel
Use cases of QuerySculptor
- AI agents discovering Pokemon types, abilities, and moves
- Exploring regional and species data with evolution chains
- Building complex GraphQL queries for any API in small steps
- Testing and validating GraphQL queries via structured tools
FAQ from QuerySculptor
What does QuerySculptor do differently from raw GraphQL query generation?
QuerySculptor provides structured MCP tools that let agents build, validate, and execute queries step-by-step rather than generating complete query strings at once.
What are the runtime dependencies?
The server uses Node.js with pnpm and is deployed on Vercel. Clone the repo and install dependencies with pnpm install.
Where does the data live for the demo?
The demo is pre-configured to use the public Pokemon API (graphql-pokeapi.vercel.app). No local database is required.
Is there a public demo I can try?
Yes, the MCP endpoint is https://agent-query-builder-toolbox.vercel.app/mcp and works with the Pokemon API out of the box.
What transports does QuerySculptor use?
It uses HTTP via MCP, exposed as a URL endpoint (e.g., /mcp). No other transports or authentication are mentioned in the README.
