Captain Cool — Cric-Oracle Pro

A lightweight FastAPI web app that simulates and (optionally) sources live IPL match scores, provides a live "radar" UI, and runs a small multi-agent strategy pipeline driven by an LLM. The UI mirrors a Captain/Analyst/Architect decision interface for in-play tactical recommendations.

This repository contains a local development demo intended for experimentation and UX prototyping.

Key features

• Real-time live radar UI served at / (WebSocket updates from /ws/live-radar).
• Toggle to attempt live-score scraping from Cricbuzz (best-effort) via /api/cricbuzz/enable.
• Multi-agent strategy endpoint /api/strategize that uses Google GenAI (Gemini) to produce Analyst, Architect, and Captain outputs.
• Several UI action endpoints: /api/update-state, /api/debate, /api/deploy, /api/impact.
• Clean, responsive UI in templates/index.html with a left live-entry panel and large hero action area.

Files of interest

• main.py — FastAPI app, WebSocket server, endpoints, and live-state simulation.
• cricbuzz.py — Best-effort scraper that tries mobile/desktop Cricbuzz and ESPN fallback for score snippets.
• templates/index.html — Front-end UI template. Buttons are wired to the backend.
• requirements.txt — Python dependencies.

Quick start (development)

1. Create a Python virtual environment and install dependencies:

python -m venv .venv
.\.venv\Scripts\activate   # Windows
pip install -r requirements.txt

2. (Optional) Set environment variables. The app expects a Gemini API key for the AI pipeline (optional). Create a .env file or set GEMINI_API_KEY in the environment:

GEMINI_API_KEY=YOUR_GEMINI_API_KEY

3. Run the app:

python -m uvicorn main:app --host 127.0.0.1 --port 8000 --reload

4. Open the UI at: http://127.0.0.1:8000

Useful API endpoints

• GET /api/live-radar — current LIVE_STATE JSON.
• POST /api/cricbuzz/enable — body { "enabled": true } to enable Cricbuzz scraping.
• GET /api/cricbuzz/status — check whether scraping is enabled.
• POST /api/strategize — run the LLM multi-agent pipeline (expects a MatchState JSON body).
• POST /api/update-state — update the shared live state with UI values.
• POST /api/debate, /api/deploy, /api/impact — UI action stubs that return confirmations.
• WebSocket: ws://127.0.0.1:8000/ws/live-radar — subscribe to live-state updates.

Notes on live scores and scraping

• The built-in cricbuzz.py scraper is best-effort and uses simple HTML/meta parsing. It may fail if Cricbuzz changes markup or blocks requests.
• For production-accurate live scores, use a dedicated sports data API (recommended). I can add adapters for providers like CricAPI, SportRadar, or similar if you provide credentials.

Troubleshooting

• If the UI cannot connect to the WebSocket, ensure the uvicorn server is running and port 8000 is available.
• If GEMINI_API_KEY is not set, the /api/strategize endpoint will fail — either set the key or avoid the AI button.
• If scraping fails frequently, enable the toggle only for testing or switch to a paid API.

Development & Contributing

• The project has been initialized as a Git repository and recent changes were pushed to the remote you provided during work.
• Create feature branches, run the server locally, and open a pull request.

License

This repository is provided as-is for prototyping and development. No license specified.