magic-workspace-init

Initialize a MAGIC data processing workspace: directory scaffolding, Python environment verification, dependency installation, and LLM configuration. Use when starting a new data project or setting up the MAGIC environment for the first time.

When It Activates

Use this skill when setting up a new workspace or installing dependencies. Trigger phrases: setup, initialize, create workspace, install dependencies, set up environment, init project.

• Starting a new data processing project
• Setting up environment for MAGIC data skills for the first time
• User asks to "set up", "initialize", "bootstrap", or "install" the data workspace
• User needs help installing Python packages or DataDesigner
• Need to verify the environment is ready for data processing

When NOT to Use: Workspace already exists and is initialized. Tier 1 quick tasks (e.g., "clean these nulls") do not need full workspace scaffolding.

Quick Facts

Tags

workspace setup init environment bootstrap install

Workspace Directory Structure

workspace/
├── data/
│   ├── input/          # Original input files (never modified)
│   ├── checkpoints/    # Intermediate results (ckpt_NN_*.csv)
│   └── output/         # Final processed data
├── logs/               # Profiling results, validation reports, analysis_journal.md
├── reports/            # Generated reports (Markdown)
├── charts/             # Generated visualizations (PNG, SVG)
└── configs/            # Agent configs, synthesis configs

For Tier 2/3 projects, the workspace also contains:

workspace/
├── workspace_state.md          # Phase tracker, quality score, task list
├── specs/
│   └── data-spec.md            # Single source of truth for dataset properties
└── logs/
    └── analysis_journal.md     # Decision log (timestamp, context, options, chosen, rationale)

Project Types

Environment Setup

Preferred method (uv — fast, no system pollution):

uv venv .venv --python 3.12
source .venv/bin/activate
uv pip install -r requirements.txt

Standard venv:

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Default Output Paths by Skill

Scripts

No scripts — workspace initialization is performed by the agent directly.

Dependencies

No dependencies — initialization only requires Python stdlib.