fal

#!/bin/bash # Run the fal.ai MCP server with Streamable HTTP transport # Set the transport mode to HTTP (uses streamable-http in main.py) export MCP_TRANSPORT=http # Set the port (default 6274) export PORT=${PORT:-6274} # Load .env file if it exists if [ -f .env ]; then export $(cat .env | grep -v '^#' | xargs) fi # Set the FAL API key if provided as argument (overrides .env) if [ -n "$1" ]; then export FAL_KEY="$1" fi echo "Starting fal.ai MCP server..." echo "" echo "Server endpoint: http://127.0.0.1:$PORT/mcp/" echo "" echo "To connect from your LLM IDE, add this to your MCP settings:" echo ' "Fal": {' echo ' "url": "http://127.0.0.1:'$PORT'/mcp/"' echo ' }' echo "" echo "Note: This is an MCP API endpoint, not a web page. Connect via your IDE." echo "" # Run the server uv run python main.py