Gel Database
About
Integrates with Gel databases to enable automated schema discovery, query validation, and execution of EdgeQL and TypeScript queries for database interactions and management tasks.
Details
- Author
- christian561
- Repository
- christian561/gel-mcp-server
- GitHub stars
- 8
- Downloads
- 336
- Categories
- Productivity, Developer Tools, Design, AI, Search, Frontend, Project Management, Infrastructure, Database
- Tags
- #integration
Jump to
- describe-schema: learn entity types, properties, relationships, and constraints.
- validate-query: validate raw EdgeQL syntax without executing.
- execute-edgeql: run raw EdgeQL queries and retrieve data.
- search-gel-docs: search Gel documentation for syntax, features, or examples.
- execute-typescript: run TypeScript Gel queries using the query builder syntax.
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Gel DatabaseCommand (node, npx, python, etc.)nodeArguments-
Argument 1
your/full/path/to/build/index.js
Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
-
Argument 1
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install dependencies (yarn install), copy your dbschema folder, initialize a Gel project (npx gel project init), generate EdgeQL JavaScript query builder (npx @gel/generate edgeql-js), update connection settings in src/index_gel.ts, build the project (yarn build), and optionally include the gel_llm.txt documentation file. Then configure the server in your MCP client (e.g., Cursor) by adding a command server pointing to node build/index.js.
describe-schema
Helps your LLM agent learn and understand your database structure without having to manually inspect the code. The agent can discover available entity types, their properties, relationships, and constraints to generate more accurate queries.
validate-query
Helps your LLM agent verify raw EdgeQL query syntax without executing it, allowing safe validation of generated queries before they're run against your database.
execute-edgeql
Helps your LLM agent directly interact with your database by running raw EdgeQL queries, retrieving data, and performing operations based on your instructions.
search-gel-docs
Allows your LLM agent to search through the Gel documentation to find relevant information about EdgeQL syntax, features, or examples, returning comprehensive results with context.
execute-typescript
Similar to execute-edgeql but can be used for testing and running Typescript Gel queries made with the query builder syntax.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"gel database": {
"env": {},
"args": [
"your/full/path/to/build/index.js"
],
"shell": false,
"command": "node"
}
}
}
Linux
{
"env": [],
"args": [
"your/full/path/to/build/index.js"
],
"shell": false,
"command": "node"
}
Macos
{
"env": [],
"args": [
"your/full/path/to/build/index.js"
],
"shell": false,
"command": "node"
}
Windows
{
"env": [],
"args": [
"/c",
"node your/full/path/to/build/index.js"
],
"shell": false,
"command": "cmd"
}
Gel Database MCP Server [Unofficial/Old]
This was created before Gel had an official MCP server which is a lot easier to setup. Leaving this one up though since it was an example of an MCP server I made. Official one here: https://github.com/geldata/gel-mcpA TypeScript-based Model Context Protocol (MCP) server designed to streamline Gel database operations with EdgeQL queries. This project provides Tools for LLM Agents (Cursor Agent, Claude Code, etc) to automate learning about your schema, and writing, validating, and executing database queries. Easily interact with your Gel database through natural language. Vibe coders rejoice!
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.





