An AI-powered game show where entrepreneurs pitch to simulated judges. Projects are researched by AI, scored by four judge personas, voted on by the community, then turned into episodes.
AI judging system for hackathon competitions. Submissions arrive through a web dashboard, get researched, scored by four AI judges, voted on with Solana tokens, then turned into episodes.
→ hackathon/README.md for the full pipeline and CLI reference
Quick start:
uv sync
clanktank --helpLive dashboard: m3org.com/tv/hackathon
Full production pipeline: Tally/Typeform → Google Sheets → SQLite → AI research → script generation → PlayCanvas rendering → video recording → YouTube.
# Import submissions from Google Sheets
python scripts/sheet_processor.py -s "Block Tank Pitch Submission" -o ./data -j --db-file pitches.db
# Research and manage pitches
python scripts/pitch_manager.py --db-file data/pitches.db --list --filter-status submitted
python scripts/pitch_manager.py --db-file data/pitches.db --research <id>
# Record an episode
node scripts/shmotime-recorder.js <episode-url>
# Upload to YouTube
python scripts/upload_to_youtube.py --from-json metadata.json aimarc |
 aishaw |
 spartan |
 peepo |
 Eliza (host) |
# Clone and install
git clone https://github.com/m3org/clanktank
cd clanktank
uv sync
# Configure environment
cp .env.example .env
clanktank config # shows what's missing
clanktank config --setup # interactive setupRequired env vars: OPENROUTER_API_KEY, DISCORD_CLIENT_ID, DISCORD_CLIENT_SECRET, DISCORD_TOKEN, PRIZE_WALLET_ADDRESS
- Backend: Python, FastAPI, SQLite, SQLAlchemy, Pydantic
- Frontend: React 18, TypeScript, Vite, Tailwind CSS
- AI: OpenRouter (judge scoring + research), Anthropic Claude (scripts), Perplexity (research)
- Integrations: ElevenLabs (voice), Discord.py (voting), Google APIs (Sheets, YouTube), Solana (voting)
- Recording: Puppeteer-based video capture of PlayCanvas-rendered episodes
AI judge decisions are simulated and do not constitute real investment advice.