Ghostshell

关于

Ghostshell functions as an MCP (Model Context Protocol) server, allowing AI assistants like Claude Desktop and Cursor to perform sophisticated causal research and conjoint analysis. It leverages synthetic populations, generated from US Census microdata, to construct statistically valid experiments for understanding human decision-making. The server provides a comprehensive suite of tools for validating research questions, generating experiment attributes, creating and managing experiments, and retrieving detailed analytical insights, accessible via both MCP clients and a direct REST API for seamless integration.

主要功能

• Synthetic Populations for Representative Sampling
• Conjoint Analysis (AMCE) for Attribute Importance
• MCP Protocol Compatibility with AI Assistants
• Causal Research & Experiment Validation
• REST API for Direct Integrations
• 2 GitHub stars

使用案例

• Validating causal research questions for AI-driven studies.
• Generating AI-powered attributes and levels for market research experiments.
• Conducting and analyzing conjoint experiments with synthetic respondents to understand decision-making.