rbutinar

rbutinar / ade-core

Public

Agentic Data Engineering Framework - Enable autonomous data workflows with AI agents

15
4
100% credibility
Found Feb 26, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

ADE Core extracts metadata like notebooks and jobs from Databricks workspaces, stores it for easy access, and lets AI tools like Claude search, analyze, and trace data dependencies.

How It Works

1
📖 Discover ADE

You hear about ADE, a helpful tool that gathers info from your data notebooks and jobs so your AI assistant can understand your projects.

2
🛠️ Start with sample data

Download the tool and launch it with built-in sample data to see how it works right away.

3
🔗 Link to your AI chat

Connect the tool to your AI helper like Claude, so it can access the gathered info seamlessly.

4
💬 Ask about your data

Chat naturally with your AI, asking things like 'What notebooks exist?' or 'Show me the sales one,' and get clear answers instantly.

5
📥 Bring in your real data

Point the tool at your data workspace to pull in your actual notebooks and jobs automatically.

6
🔍 Explore and trace connections

Search for any data item, view details, or see how notebooks connect to tables, all through your AI chat.

🎉 AI becomes your data partner

Your AI now fully understands your data setup, helping you analyze, trace, and manage everything effortlessly.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 15 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is ade-core?

ADE Core is a Python framework for agentic data engineering that extracts metadata from Databricks notebooks, jobs, and source code, exposing it as a searchable catalog for AI agents via the Model Context Protocol (MCP). It tackles scattered data contexts—like SQL in views or PySpark in notebooks—by letting you query assets naturally in Claude Code or Desktop: "What tables does this notebook write to?" or "Show sales aggregation code." Quick CLI extraction and demo data get you querying agentic data analysis in minutes.

Why is it gaining traction?

It bridges AI agents to real data platforms without custom tooling, using MCP for seamless Claude integration that feels like agentic GitHub Copilot for data. The lineage tracing and platform stats stand out for agentic data management, while synthetic demo data hooks devs testing agentic data science fast. Expansions to Power BI and PostgreSQL signal broader agentic data plane potential.

Who should use this?

Databricks data engineers debugging pipelines via AI queries. Agentic data scientists tracing notebook dependencies without manual hunts. Teams prototyping agentic data specialist workflows in Salesforce-like environments or agentic databases.

Verdict

With 14 stars and 1.0% credibility, this 0.1.0-alpha is immature—light docs, no tests—but the MCP demo delivers instant value for agentic data engineering experiments. Prototype with it now; hold for production until more platforms ship.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.