ChatSpatial

MCP server for spatial transcriptomics analysis via natural language.

Analyze your spatial data from any MCP-compatible client. No coding required.

ChatSpatial exposes 20 schema-validated MCP tools. Those tools orchestrate 65 spatial transcriptomics methods across 15 analytical categories; the tools are the user-facing interface, and the methods are selected through tool parameters.

20 MCP tools orchestrating 65 methods across 15 analytical categories. Supports 10x Visium, Xenium, Slide-seq, MERFISH, and more.


Choose Your Path

New to ChatSpatial:

InstallationConfiguration GuideQuick Start

Already running ChatSpatial:

Concepts for method selection, Examples for prompt recipes, and Methods Reference for exact parameters.

Something failed:

Start with Troubleshooting, then use Frequently Asked Questions for short answers and pointers.

Contributing:

Use Contributing to find the right contribution path.


Getting Started

Installation

Set up the environment

Install ChatSpatial and prepare your Python environment.

Installation
Quick Start

Your first analysis

Load data and start analyzing with natural language.

Quick Start
Docker / GHCR

Run without local dependency resolution

Pull the GHCR image, mount data, and configure MCP.

Docker / GHCR
Concepts

Understand the methods

When to use which method and why.

Concepts
Examples

See what’s possible

Natural language commands for every analysis type.

Examples

Reference

Methods Reference

MCP tools, supported methods, parameters, and defaults

Methods Reference
Configuration

MCP client configuration

Configuration Guide
Data Management

External dataset registry and fetch workflow

Data Management
Contributing

Contributor entry point for docs and code changes

Contributing

Support