Multi-Agent Investment Intelligence Platform — Built for the Swarms ACM Hackathon 2026
Four AI agents: Security Auditor, Quant, Sentiment Analyst, and Lead Partner - debate the investment thesis of any cryptocurrency token in real-time.
The Virtual VC Boardroom simulates a hedge fund investment committee meeting. When a user inputs any token or protocol:
- 🔒 Security Auditor — Hunts for red flags: rug-pull risks, contract vulnerabilities, centralization vectors
- 📊 The Quant — Analyzes tokenomics: FDV, MC/FDV ratio, inflation rates, on-chain metrics, fair value models
- 🐦 Sentiment Analyst — Scans X/Twitter narrative: real hype vs. fake hype, influencer consensus, community health
- 🏛️ Lead Partner — Synthesizes all reports and delivers a GO / NO-GO / HOLD verdict with recommended portfolio allocation
- Multi-Agent Debate System: Not just Q&A — agents debate and challenge each other's findings
- Frenzy Mode: Toggle to activate a simulated bonding curve trading terminal for the agent's own token (
$VCB) - Production-Ready Backend: Built on the official
swarmsPython framework with FastAPI - Stunning UI: Glassmorphic dark dashboard with soundwave animations, typing indicators, and dynamic glows
cd frontend
npm install
npm run devThe dashboard launches at http://localhost:5173 with full mock mode — no API keys needed.
cd backend
pip install -r requirements.txt
# Set your API key
set OPENAI_API_KEY=your-key-here
# Run the server
python main.pyThe API launches at http://localhost:8000 with Swagger docs at /docs.
python agents.py SOL
python agents.py JUP
python agents.py WIFToggle the FRENZY switch in the top-right to activate:
- 📈 Bonding Curve Chart — Real-time price visualization
- 📖 Order Book — Simulated bid/ask spread
- ⚡ Live Trade Feed — Streaming buy/sell transactions
- 📊 Migration Progress — Progress toward DEX listing threshold
This demonstrates how the agent would function as a tokenized asset on the Swarms Marketplace.
swarms/
├── frontend/ # React + TypeScript + Vite
│ ├── src/
│ │ ├── components/ # BoardroomTable, DebatePanel, VerdictCard, FrenzyTerminal
│ │ ├── data/ # Mock debate scripts (SOL, JUP, WIF)
│ │ ├── utils/ # Frenzy mode bonding curve simulator
│ │ └── App.tsx # Main orchestrator
│ └── package.json
├── backend/ # Python + Swarms SDK + FastAPI
│ ├── agents.py # 4 Agent definitions with system prompts
│ ├── main.py # FastAPI server
│ └── requirements.txt
└── README.md
| Requirement | Status |
|---|---|
| Build a high-quality agent for real-world use in DeFi/Finance | ✅ Multi-agent investment committee |
| Enable Frenzy Mode | ✅ Full bonding curve + trade terminal |
| Publish on Swarms Marketplace | ✅ Published on Swarms World |
| Tokenize and list for sale | ✅ $VCB token is live & trading |
| Real-world utility | ✅ Investment analysis across any Solana token |
- Frontend: React 19, TypeScript, Vite, Framer Motion, Recharts
- Backend: Python 3.13, Swarms SDK, FastAPI, Uvicorn
- Chain: Solana
- Agent Framework: Swarms multi-agent orchestration
MIT — Built for the Swarms ACM Hackathon 2026