home / mcp / fatebook mcp server
Your new, favorite MCP server for what should be your favorite personal predictions tracker, fatebook.io.
Configuration
View docs{
"mcpServers": {
"an1lam-fatebook-mcp": {
"command": "uvx",
"args": [
"fatebook-mcp"
],
"env": {
"FATEBOOK_API_KEY": "your-api-key-here"
}
}
}
}You can run Fatebook as an MCP server to integrate prediction tracking with your AI assistants. This server lets you create, manage, and observe Fatebook predictions directly through MCP, so you can build interactive workflows with Claude or other MCP clients.
Connect to the Fatebook MCP server from your MCP client. Use a local stdio configuration to run the server as a child process and communicate through standard input/output.
Prerequisites you need before installing the Fatebook MCP server.
- Python 3.13 or higher.
- uv or pip for managing MCP servers.
- A Fatebook account with an API key.
Install options
# Install from PyPI (recommended)
pip install fatebook-mcp
# Or add using uv (installs MCP engine if needed)
uv add fatebook-mcpYou will need to provide your Fatebook API key to access the Fatebook account.
Environment variable to set in your MCP client configuration.
Create a new prediction with a forecast probability.
List existing predictions with optional filters to narrow results.
Retrieve full information about a specific prediction.
Update the predicted probability for an existing question.
Attach notes or reasoning to a prediction.
Mark a prediction as YES, NO, or AMBIGUOUS to indicate its outcome.
Update titles, resolve dates, and notes for existing questions.
Remove questions you no longer need.
Track your forecasting activity across questions.