Installation

This page describes the recommended installation path for CellPainting-Claw.

The goal of installation is to obtain a working installation of the main package, the public CLI commands, and the optional OpenClaw-facing MCP support.

Prerequisites

Before installing the package, make sure you have:

  • a local checkout of the CellPainting-Claw repository

  • Conda available on the target machine

  • permission to create a dedicated environment for the toolkit

CellPainting-Claw currently ships with a validated Conda environment file:

  • environment/cellpainting-claw.environment.yml

That environment targets Python 3.10 and includes the main toolkit dependencies used by this repository.

Available Components After Installation

After installation, the main things you can use are:

  • the Python packages cellpainting_claw and cellpainting_skills

  • the CLI commands cellpainting-claw and cellpainting-skills

  • the optional MCP support used by OpenClaw

Optional Extras

The project also defines optional extras in pyproject.toml. These are mainly useful for a more customized environment rather than the standard installation path.

pip install -e .[test]
pip install -e .[mcp]
pip install -e .[data-access]
pip install -e .[deepprofiler]

In the current repository, the validated Conda environment already includes the main toolkit stack used by the standard installation path.

Verify The Installation

A minimal post-installation check is to confirm that the public CLI entrypoints are available:

cellpainting-claw --help
cellpainting-skills --help

You can also verify the import surface in Python:

import cellpainting_claw
import cellpainting_skills

If both the CLI and imports work, the package layer is installed correctly.