🏏 Captain Cool - Multi-Agent IPL Strategist 🏏

Welcome to Captain Cool, a robust, multi-agent AI orchestration pipeline designed to simulate the elite decision-making process of legendary IPL T20 cricket captains. Powered by the Gemini 2.5 Flash model and the official Google GenAI SDK, this tool acts as a comprehensive match strategist for the IPL 2026 season.

Overview

Captain Cool utilizes a three-agent architecture to evaluate real-time IPL match situations, fetch relevant statistics, propose tactical moves, aggressively challenge them, and make a finalized "Captain Cool" decision.

The Agents

1. Stats Analyst 📊: Consumes the match state and leverages tool calling (fetch_live_match_data, fetch_historical_stats) to fetch precise numbers, matchups, pitch conditions, and historical data.
2. Strategist (Initial Proposal) 🧠: Acting as "Captain Cool," this agent reviews the Analyst's report and the current match state to formulate a single, actionable tactical plan using authentic cricket terminology.
3. Devil's Advocate 👿: A highly critical assistant coach who aggressively challenges the Strategist's proposal. This agent looks for counter-matchups, momentum risks, and flaws in the plan.
4. Strategist (Final Decision) 🏆: The Captain's final word. After absorbing the Devil's Advocate's critique, the Strategist either pivots or sticks to the original plan, explaining the cricket logic behind the decision.

Features

• Interactive CLI: A beautiful command-line interface powered by rich, prompting users for real-time match details (Innings, Required Run Rate, Pitch Conditions, Bowlers Remaining, etc.).
• Tool Calling: The Stats Analyst automatically calls mock Python functions to fetch live data and historical player statistics.
• Resiliency: Incorporates the tenacity library for exponential backoff retries, ensuring stability against API rate limits.
• Authentic Tone: The prompt engineering ensures agents speak in true cricket jargon without generic AI terminology.

Setup & Installation

1. Clone the Repository: Ensure you have the captain_cool directory.
2. Install Dependencies:

   pip install -r requirements.txt

   (Requires google-genai, rich, tenacity, python-dotenv)
3. Environment Variables: Create a .env file in the root directory and add your Gemini API Key:

   GEMINI_API_KEY=your_api_key_here

Usage

Run the main script to start the interactive CLI:

python main.py

You will be prompted to enter match details (or press Enter to use defaults). The pipeline will then kick off, and you'll see the multi-turn debate unfold in your terminal with colored panels.

To run a programmatic test without user input:

python test_run.py

Project Structure

• main.py: The entry point for the interactive CLI application.
• orchestrator.py: Contains the multi-agent pipeline logic, handling the turn-by-turn conversation and API calls.
• agents.py: Holds the system prompts defining the persona of each agent.
• tools.py: Contains the tool functions (fetch_historical_stats, fetch_live_match_data) used by the Stats Analyst.
• test_run.py: A script for running programmatic tests of the pipeline.

Technologies Used

• Google GenAI SDK (google-genai)
• Gemini 2.5 Flash
• Python 3.10+
• Rich (for terminal UI)
• Tenacity (for robust API retries)