Forecasting MCP

This is a Model Context Protocol (MCP) implementation for interacting with a forecasting website. It enables LLMs to make, update, and delete forecasts, as well as retrieve forecast data and access external information through Perplexity API.

Features

• Make forecasts on an existing forecasting website
• Update existing forecasts
• Get information about forecasts
• Access Perplexity AI for additional context and research

Setup

1. Install the required dependencies:

pip install .

2. Create a .env file with the following variables:

API_URL=<your_forecasting_api_url>
BOT_USERNAME=<your_bot_username>
BOT_PASSWORD=<your_bot_password>
PERPLEXITY_API_KEY=<your_perplexity_api_key>  # For accessing Perplexity API

Usage

Run the server with:

uv run forecasting.py

Available MCP Tools

Forecasts Management

• get_forecasts: List forecasts by category and status
• get_forecast_data: Get details for a specific forecast
• update_forecast: Create a new forecast point for a forecast.

Forecast Points

• get_forecast_points: Get all forecast points for a forecast

External Information

• query_perplexity: Query the Perplexity API for additional information to inform forecasts

Dependencies

• Python 3.13+
• FastAPI
• httpx
• MCP
• requests

Development

This server uses FastAPI and the MCP library to provide a standardized interface for LLMs to interact with forecasting data.