Installation

Installation

Prohori is on PyPI: https://pypi.org/project/prohori/. The latest release is 0.1.0.

Requirements

• Python 3.11, 3.12, 3.13 or 3.14
• Linux or Windows (tested in CI); macOS should work but is not tested
• uv (recommended) or pip

From PyPI (recommended)

pipx install prohori        # the prohori command in its own environment
prohori demo

or, inside a virtual environment:

pip install prohori
prohori --version

From source

Use this to work on Prohori or to run the commands that use the sample files in the repository.

git clone https://github.com/UtshaBasak/Prohori.git
cd Prohori
uv sync
uv run prohori --version

With plain pip:

python -m venv .venv
# Linux/macOS: source .venv/bin/activate
# Windows:     .venv\Scripts\activate
pip install -e .
prohori --version

Optional extras

The base install is deliberately small. Heavy dependencies are opt-in:

Extra	Adds	Install
ml	Machine-learning detection on network flows and log features	pip install "prohori[ml]"
pipeline	Streaming pipeline and warehouse clients	pip install "prohori[pipeline]"
llm	Optional LLM incident analyst	pip install "prohori[llm]"
all	All of the above	pip install "prohori[all]"

The extras are declared now and filled in as each feature is released.

Configuration and API keys

Every external service is optional; Prohori works fully offline. To use enrichment or the LLM analyst, copy .env.example to .env and fill in only the keys you need. .env is git-ignored and must never be committed.