Installation¶

Requirements¶

• Python 3.11 or newer
• enough local disk space for the requested output
• a dedicated working directory for inputs and generated artifacts

Install From PyPI¶

Create an isolated environment:

macOS and Linux¶

python3 -m venv .venv
source .venv/bin/activate
python3 -m pip install --upgrade pip
python3 -m pip install agent-paranoid-android

The base package supports CSV and JSON workflows. Add only the integrations you need:

python3 -m pip install "agent-paranoid-android[parquet]"
python3 -m pip install "agent-paranoid-android[mcp]"
python3 -m pip install "agent-paranoid-android[mcp,trino]"
python3 -m pip install "agent-paranoid-android[all]"

Use test-data-agent doctor --require-extra all to verify a full installation.

Windows PowerShell¶

py -3.11 -m venv .venv
.venv\Scripts\Activate.ps1
python -m pip install --upgrade pip
python -m pip install agent-paranoid-android

Run the self-contained environment check:

test-data-agent doctor

Expected final lines:

quickstart smoke: ok
doctor passed

doctor creates its sample input and output under a temporary directory. It does not need a repository checkout and does not contact Trino or another external system.

Install For Development¶

Clone the repository and install the locked development environment:

git clone https://github.com/wa-pis/agent-paranoid-android.git
cd agent-paranoid-android
python3 -m pip install "uv==0.11.23"
uv sync --frozen --all-extras --no-install-project
uv sync --frozen --all-extras --no-editable --no-build-isolation

Run the release-quality checks:

uv run --no-sync scripts/check_release.sh

Build the documentation in its isolated dependency group:

uv sync --frozen --only-group docs --no-install-project
uv run --no-sync mkdocs build --strict

Confirm The Installed Version¶

python3 -c "import test_data_agent; print(test_data_agent.__version__)"

The package name is agent-paranoid-android. The command remains test-data-agent.

Next Step¶

Use First CSV Dataset for one table or Related Tables for a folder containing one CSV per table.