Jupyters - AI Notebook Assistant | Give Your AI Superpowers

Everything Your AI Needs for Notebooks

16 powerful MCP tools for complete Jupyter integration

📓

Read & Write

Full access to notebook structure. Read cells, update content, add new cells, merge, split.

⚡

Live Execution

Execute code in real Jupyter kernels. Capture outputs, stdout, stderr, and errors.

🔍

Variable Inspection

Deep semantic understanding of DataFrames, arrays, tensors. See shapes, schemas, samples.

Pro

📊

Plot Capture

See matplotlib, seaborn, and plotly visualizations directly in AI responses.

Pro

🧠

Auto Error Analysis

When code fails, Jupyters automatically inspects variables in the traceback for context.

🛡️

Safety Checks

Prevent accidental file deletion, network calls, and other destructive operations.

🎯

Domain Profiles

Enhanced inspection for ML models (PyTorch, TensorFlow) and financial data (OHLCV).

Team

💰

Token Budgeting

Smart truncation prevents overwhelming your AI with massive outputs.

See It In Action

Real conversations with Claude using Jupyters

Load sales_data.csv into my notebook and show me the top 10 products by revenue

I'll create a cell to load and analyze your sales data.

import pandas as pd
df = pd.read_csv('sales_data.csv')
top_products = df.groupby('product')['revenue'].sum().nlargest(10)
top_products

Executed! Here are your top 10 products:

Product A	$1,234,567
Product B	$987,654
Product C	$876,543

Check my model training notebook - what's the current accuracy and can you suggest improvements?

Let me inspect your model and training results...

🔧 Using: inspect_variable('model')

Model Architecture: ResNet18 with 11.7M parameters

Current Accuracy: 78.3% (validation)

Suggestions:

Learning rate seems high (0.01) - try 0.001
Add data augmentation for better generalization
Consider learning rate scheduling

Cell 5 is throwing a KeyError. Can you fix it?

Let me run cell 5 and analyze the error...

KeyError: 'total_sales'

Auto-Analysis: The DataFrame doesn't have a 'total_sales' column. Available columns are: ['date', 'product', 'quantity', 'price']

I've fixed the cell to calculate total_sales first:

df['total_sales'] = df['quantity'] * df['price']
monthly = df.groupby('date')['total_sales'].sum()

Now it works! ✓

Get Started in 2 Minutes

Three simple steps to give your AI notebook superpowers

Install

pip install jupyters-server

Configure

Add to Claude Desktop config:

{
  "mcpServers": {
    "jupyters": {
      "command": "jupyters-server"
    }
  }
}

Config location:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

Use It!

Restart Claude Desktop, then try:

"Read my notebook at /path/to/notebook.ipynb and summarize it"

Need unlimited executions?

Free tier includes 10 executions per day. Upgrade to Pro for unlimited.

See Pricing

Give Your AI Superpowers with Jupyter Notebooks