What runtime and dependencies are required?
Python 3.11+ is required for local execution. Dependencies are listed in `requirements.txt` and installed via pip.
@subconscious-ai
Basic information
Run conjoint experiments and causal research through AI-powered behavioral simulations
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"subconscious-ai": {
"url": "https://ghostshell-runi.vercel.app/api/sse?token=YOUR_TOKEN"
}
}
}Tool list not available yet
This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.
This server requires authentication. Tool discovery may require signing in or credentials from the provider.
Fetch the live tool list directly from this server's endpoint using the button above.
Subconscious Ai Mcp is an MCP (Model Context Protocol) server that lets you run AI-powered conjoint experiments from Claude Desktop, Cursor, or any MCP-compatible client. It uses causal inference and synthetic populations based on US Census microdata to help understand why people make decisions.
You can use the hosted server by adding a URL with a token to your MCP client configuration (Claude Desktop or Cursor), or run it locally by cloning the repo, installing dependencies with Python 3.11+, and setting environment variables AUTH0_JWT_TOKEN and API_BASE_URL. Tokens are obtained from app.subconscious.ai.
Python 3.11+ is required for local execution. Dependencies are listed in `requirements.txt` and installed via pip.
You need an active Subconscious AI subscription. For the hosted server, include your token as a query parameter in the SSE URL. For local setup, set the `AUTH0_JWT_TOKEN` environment variable.
The server proxies requests to the Subconscious AI backend (`https://api.subconscious.ai`). Self-hosted deployments also route through this backend; user data is not stored on the MCP server itself.
The server supports MCP via SSE (Server-Sent Events) and also provides a REST API at `/api/call/{tool}` endpoints. Both require authentication.
The README mentions experiments can be created with a specified confidence level (e.g., "Reasonable"), but does not state explicit quantitative limits. The platform surveys synthetic respondents (e.g., 500 per experiment).

See and control a real Windows PC you own — from any MCP client, locally or remotely. Observe (UIA + screenshots), click/type/drag/scroll, launch apps, owner Live View. BYOH: your machine, your key.

Yuppsy MCP: build and publish real estate websites from your AI assistant. Yuppsy lets your AI assistant create live websites for real estate agents and individual property listings. You don't need a designer or any code
Senso MCP turns a Joomla 5 or 6 site into a remote MCP server. It runs inside Joomla itself (no extra server, no Node.js bridge) and is secured with OAuth 2.1 and Joomla's own ACL. The free pack lets an AI assistant crea
Hindsight: Agent Memory That Learns
A one-of-a-kind resume builder that keeps your privacy in mind. Completely secure, customizable, portable, open-source and free forever. Try it out today!
Comments