Intugle Mcp

Name: Intugle Mcp
Author: Intugle

@Intugle

代码仓库

关于 Intugle Mcp

Generate automated semantic models using data engineering agents and built data products on demand

#semantic-model #data-product #semantic-layer #data-analysis

基本信息

分类

数据与分析

传输方式

stdio

发布者

Intugle

提交者

raphael-intugle

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "data-tools": {
      "command": "python",
      "args": [
        "-m",
        "venv",
        ".venv"
      ]
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Intugle Mcp?

Intugle Mcp is a GenAI-powered open-source Python library that builds a semantic data model over existing data systems. It profiles, classifies, and discovers links and relationships across data assets, enriches them with business glossaries, and enables semantic search and auto-generated queries for unified data products.

How to use Intugle Mcp?

Install the package with pip install intugle in a virtual environment. Configure an LLM provider via environment variables (LLM_PROVIDER, API_KEY) and optionally a Qdrant vector database for semantic search. Use the SemanticModel class to profile, link, and glossarize datasets, then use DataProduct to generate unified data products.

Key features of Intugle Mcp

Semantic data model that transforms fragmented datasets into a connected graph.
Auto-generated business glossary and semantic search across technical and business users.
Data product generation with auto-joined SQL queries and reusable outputs.

Use cases of Intugle Mcp

Data engineers automate profiling, classification, and linking of fragmented data assets.
Data analysts and scientists accelerate data readiness with contextual intelligence and auto-generated SQL.
Business analysts query data via natural language semantic search without technical dependencies.
Organizations unify data from heterogeneous sources (CSV, Snowflake, Databricks) into consistent data products.

FAQ from Intugle Mcp

How do I configure an LLM for Intugle Mcp?

Set the LLM_PROVIDER environment variable (e.g., openai:gpt-3.5-turbo) and the corresponding API key (e.g., OPENAI_API_KEY).

What are the runtime requirements?

Python with a virtual environment. On macOS, you may need to install libomp via Homebrew and run SSL certificates command if using official Python installer.

Does Intugle Mcp work with Databricks or Snowflake?

Yes. Quickstart notebooks are provided for Databricks Unity Catalog, Snowflake Horizon Catalog, and native Snowflake with Cortex Analyst, as well as native Databricks with AI/BI Genie.

What license is Intugle Mcp released under?

Apache 2.0.

How do I enable semantic search?

You need a running Qdrant instance (Docker command provided) and set QDRANT_URL and optionally QDRANT_API_KEY. Only OpenAI embeddings are currently supported for semantic search.