Portfolio-aware perpetual-futures trading agent using 0G Compute for trade decisions and 0G Storage for verifiable alpha delivery.

Live App: https://openclaw.maxxit.ai/ · OpenClaw Skill: clawhub.ai/abhi152003/maxxit-0g

Maxxit 0G:

1. Pulls live portfolio and market context for a user.
2. Calls 0G Compute for a portfolio-aware trade decision: should trade, side, confidence, and reasoning.
3. Executes the trade on Ostium, Aster, or Avantis perpetuals after explicit user confirmation.
4. Optionally publishes the executed trade as paid alpha, with the alpha payload stored on 0G Storage and a rootHash returned to every buyer for independent verification.

The OpenClaw skill in skills/maxxit-0g/ is the production agent interface. It ships with the repo and drives the full workflow.

Agent identity is wallet-based, and portable alpha payloads are stored through 0G Storage so the agent's outputs can be verified outside Maxxit's application database.

The agent uses 0G Compute for trade decisions:

1. Resolve the user's lazy-trading deployment + portfolio context.
2. Run market research (/api/lazy-trading/research).
3. Send { deploymentId, tokenSymbol, marketResearchSummary } to /alpha/0g-decision.
4. Call 0G Compute via @0glabs/0g-serving-broker.
5. Return { shouldTrade, side, confidence, reasoning, modelUsed, verified }.
6. Ask for confirmation before executing a trade.

Key files:

• pages/api/lazy-trading/programmatic/alpha/0g-decision.ts
• lib/zg-compute.ts
• 0g-backend/src/compute.ts - broker init, getRequestHeaders, processResponse
• 0g-backend/src/server.ts

When a producer flags a trade as alpha, the alpha payload is uploaded to 0G Storage via @0gfoundation/0g-ts-sdk (Indexer.upload(MemData)). The listing returns ogStorage.rootHash and ogStorage.txHash. Buyers receive the rootHash with their purchase receipt and can use it to verify the payload.

Key files:

• pages/api/lazy-trading/programmatic/alpha/flag.ts
• pages/api/lazy-trading/programmatic/alpha/listings.ts
• pages/api/lazy-trading/programmatic/alpha/verify.ts
• lib/zg-storage.ts, lib/alpha-content-hash.ts
• 0g-backend/src/storage.ts - uploadAlphaContent, downloadAlphaContent

• @0glabs/0g-serving-broker - 0G Compute broker
• @0gfoundation/0g-ts-sdk - 0G Storage indexer client

No custom contracts were deployed. Maxxit 0G interacts directly with the 0G Galileo Storage flow contract.

• Example storage upload tx: 0xac0df9...f913bb50
• Network: 0G Galileo testnet (evmrpc-testnet.0g.ai, chain id 16602)
• Indexer: indexer-storage-testnet-turbo.0g.ai

User via OpenClaw / Maxxit UI
  |
  v
skills/maxxit-0g
  - workflow and routing rules
  - strategy runners
  |
  v
Maxxit Lazy Trading API
  |
  +-- market research
  |
  +-- decision call
  |     |
  |     v
  |   0g-backend /compute
  |     |
  |     v
  |   0G Compute
  |
  +-- trade execution
  |     |
  |     v
  |   Ostium / Aster / Avantis
  |
  +-- alpha listing
        |
        v
      0g-backend /storage
        |
        v
      0G Storage rootHash

The 0g-backend is an Express bridge that owns the 0G wallet and exposes /compute and /storage endpoints to the main Next.js app. 0G keys do not live in the user-facing process.

• Ostium perpetuals (Arbitrum)
• Aster DEX perpetuals (BNB Chain)
• Avantis DEX perpetuals (Base)
• Maxxit Alpha Marketplace for buying/selling trade signals (x402 USDC settlement)

npm install
cp .env.example .env
npm run dev

Required env:

NEON_REST_URL=
NEON_REST_TOKEN=
ZG_BACKEND_URL=http://localhost:8787
ZG_BACKEND_AUTH_TOKEN=change-me

cd 0g-backend
npm install
cp .env.example .env
npm run dev

Required env:

PORT=8787
ZG_BACKEND_AUTH_TOKEN=change-me
ZG_WALLET_PRIVATE_KEY=
ZG_COMPUTE_RPC_URL=https://evmrpc-testnet.0g.ai
ZG_COMPUTE_PROVIDER_ADDRESS=
ZG_STORAGE_RPC_URL=https://evmrpc-testnet.0g.ai
ZG_STORAGE_INDEXER_URL=https://indexer-storage-testnet-turbo.0g.ai

npx clawhub@latest install maxxit-0g
export MAXXIT_API_KEY="lt_your_api_key_here"
export MAXXIT_API_URL="https://maxxit.ai"

The skill source is in skills/maxxit-0g/ and includes 7 standalone Python strategy runners (EMA, RSI/Bollinger, Donchian/ADX, taker-flow, mean-reversion, breakout, VWAP).

Example prompts:

Should I trade BTC?
What does 0G think about ETH?
Make a portfolio-aware decision for SOL.
Open the trade and list it as alpha on 0G storage.
Run the VWAP strategy on BTC/USD using Avantis.

After a buyer completes the x402 purchase flow, they get a receipt containing ogStorage.rootHash. They can re-fetch and verify the payload:

# Verify the purchased alpha payload.
curl -L "https://maxxit.ai/api/lazy-trading/programmatic/alpha/verify" \
  -H "X-API-KEY: $MAXXIT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "listingId": "<id>", "content": <alpha object from purchase> }'

If the payload does not match the listing, verified returns false.

maxxit-latest/
pages/api/lazy-trading/programmatic/   Trading and alpha endpoints
  alpha/
    0g-decision.ts                      0G Compute entry point
    flag.ts                             Listing and 0G Storage upload
    verify.ts                           rootHash verification
lib/
  zg-compute.ts                         Compute client wrapper
  zg-storage.ts                         Storage client wrapper
  alpha-content-hash.ts                 Canonical hashing
0g-backend/                             Express bridge that owns the 0G wallet
  src/{compute.ts, storage.ts, server.ts}
skills/maxxit-0g/                       OpenClaw agent
  SKILL.md
  *-strategy.py                         Strategy runners
docs/0G_INTEGRATION_PLAN.md