AgentVault AI

AI x Web3 hackathon MVP for the 0G APAC Hackathon. AgentVault AI accepts a wallet address or crypto project name, generates an AI intelligence report, stores the report on 0G Storage testnet, and displays persistent report history.

Stack

• Frontend: React, Vite, TailwindCSS
• Backend: Django, Django REST Framework, SQLite
• AI: OpenAI/DeepSeek/OpenAI-compatible providers, deterministic local fallback for demos
• Wallet data: Etherscan V2 indexed transaction history with JSON-RPC fallback
• Storage: 0G Storage testnet through the official Python SDK

Project Structure

agentvault-ai/
  backend/
    agentvault/
    reports/
    manage.py
    requirements.txt
    .env.example
  frontend/
    src/
    package.json
    tailwind.config.js
  README.md

Backend Setup

cd backend
python -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt
copy .env.example .env
python manage.py migrate
python manage.py runserver

Edit backend/.env:

DJANGO_SECRET_KEY=replace-me
DEBUG=True
DEEPSEEK_API_KEY=
DEEPSEEK_BASE_URL=https://api.deepseek.com
OPENAI_API_KEY=
OPENAI_BASE_URL=https://api.openai.com/v1
AI_API_KEY=
AI_BASE_URL=
AI_MODEL=
OG_PRIVATE_KEY=0x-your-funded-0g-testnet-private-key
OG_BLOCKCHAIN_RPC=https://evmrpc-testnet.0g.ai
WALLET_RPC_URL=https://evmrpc-testnet.0g.ai
WALLET_SCAN_BLOCKS=8
WALLET_SCAN_SECONDS=6
ETHERSCAN_API_KEY=
ETHERSCAN_CHAIN_ID=1
ETHERSCAN_TX_LIMIT=40
OG_INDEXER_RPC=https://indexer-storage-testnet-turbo.0g.ai
OG_EXPLORER_TX_URL=https://chainscan-galileo.0g.ai/tx/{tx_hash}

OG_PRIVATE_KEY must belong to a funded 0G testnet wallet for real storage uploads. If it is missing, /api/store/ returns a local demo storage record marked as local-demo.

Frontend Setup

cd frontend
npm install
npm run dev

Open the Vite URL, usually http://localhost:5173.

API

POST /api/analyze/

Request:

{ "query": "0x1234...abcd", "chain_id": 1 }

Response:

{
  "query": "0x1234...abcd",
  "input_type": "wallet",
  "raw_data": {},
  "ai": {
    "ai_provider": "openai",
    "ai_model": "gpt-4o-mini",
    "ai_status": "live"
  },
  "report": {
    "summary": "",
    "risk_level": "low",
    "key_behaviors": [],
    "insights": [],
    "recommendation": ""
  }
}

POST /api/store/

Request:

{
  "query": "0x1234...abcd",
  "input_type": "wallet",
  "raw_data": {},
  "report": {}
}

Response includes root_hash, tx_hash, explorer_url, and the saved report record.

GET /api/history/

Returns stored report metadata and report JSON.

GET /api/report/<id>/

Returns one stored report. If a 0G root_hash exists, the backend attempts to retrieve the report from 0G Storage and falls back to the database copy while files propagate.

Demo Flow

1. Enter a wallet address, project name, or website URL.
2. Select the wallet chain for Etherscan-indexed analysis.
3. Click Generate Report.
4. Review the AI report on the dashboard.
5. Click Store on 0G.
6. Open History to see persistent reports and blockchain proof links.

Notes for Judges

• Real 0G Storage integration lives in backend/reports/storage_0g.py.
• The upload code uses the official Python SDK package 0g-storage-sdk and testnet endpoints.
• Wallet analysis uses Etherscan V2 indexed normal, internal, and ERC-20 transfer history when ETHERSCAN_API_KEY is configured.
• The dashboard displays whether AI came from a live provider or local fallback.
• AI output is normalized and validated into the required JSON shape.
• JSON-RPC and deterministic fallback paths keep the demo usable if third-party APIs are unavailable.