Quick Start

Let's load a CSV file and get a quality report — your first data task with MAGIC Data Agent Skills.

Step 1: Set Up Your Workspace

Ask your AI assistant:

"/magic:init-workspace"

This creates the workspace directory structure with data/, logs/, and checkpoint directories.

Step 2: Load and Profile Data

Place a CSV file in your project directory (or use any accessible path), then ask:

"Load sales_data.csv and tell me about the data quality"

The agent will:

1. Activate magic-data-loading — detect the format, encoding, and delimiter
2. Load the file into a pandas DataFrame
3. Checkpoint the raw data as ckpt_01_raw_data.csv
4. Transition to magic-data-profiling — run quality scoring and distribution analysis
5. Report the quality score, issues found, and recommended next steps

Step 3: Generate a Report

"Generate a report of the profiling results"

The agent activates magic-report-generation and produces a structured Markdown report with tables, findings, and recommendations.

What's Next?