@echo off REM ============================================ REM PyCatan with AI Agents - Browser Settings Mode REM ============================================ REM Opens a setup page where the user chooses: REM - Gemini model REM - API key mode: environment keys or typed keys REM - Player count and names REM Then starts the same full-auto AI game flow. REM ============================================ set LOGLEVEL=DEBUG echo. echo ================================================================================ echo PyCatan AI System - PLAY_WITH_SET_SETTINGS echo ================================================================================ echo. echo Starting components: echo 1. AI Viewer (http://localhost:5001) - Shows AI prompts and responses echo 2. LLM Logger - Shows real-time AI communication echo 3. Browser setup page - Gemini model, API key, players echo 4. Catan Game with AI Agents (AUTO MODE) echo. echo ================================================================================ echo. cd /d "%~dp0" where python >nul 2>nul if %ERRORLEVEL% NEQ 0 ( echo ERROR: Python not found in PATH pause exit /b 1 ) if exist ".venv\Scripts\python.exe" ( set PYTHON_CMD=.venv\Scripts\python.exe echo [OK] Using virtual environment ) else ( set PYTHON_CMD=python echo [!] No virtual environment found, using system Python ) echo. echo [1/3] Starting AI Viewer... start "AI Viewer - http://localhost:5001" cmd /k "%PYTHON_CMD% examples\ai_testing\web_viewer.py" timeout /t 2 /nobreak >nul echo [OK] AI Viewer started at http://localhost:5001 echo. echo [2/3] Starting LLM Logger Console... start "LLM Logger - Communication Log" cmd /k "%PYTHON_CMD% examples\ai_testing\llm_logger_console.py" timeout /t 1 /nobreak >nul echo [OK] LLM Logger console opened echo. echo [3/3] Opening browser setup and starting game after submit... echo. echo ================================================================================ echo PLAY_WITH_SET_SETTINGS echo ================================================================================ echo A browser page will ask for model, run mode, replay/session options, players, and keys. echo Default key mode: use GEMINI_API_KEY / ELEVENLABS_* from ENV or .env when blank. echo Force typed keys: add --ask-api-keys echo Explicit ENV mode: add --use-env-keys echo Run modes include new game, resume old session, watch replay, and analyse game. echo Reactions can run async, sync/no parallel, or be disabled. echo The setup page also lets you choose English or Hebrew table talk. echo After you submit, the board will load automatically in that same tab. echo Press Ctrl+C here to stop the game. echo ================================================================================ echo. %PYTHON_CMD% examples\ai_testing\play_with_ai.py --auto --browser-settings %* echo. echo ================================================================================ echo Game session complete! echo ================================================================================ echo. echo Session logs saved to: examples\ai_testing\my_games\ echo. pause