Thetanuts

When no wallet is configured, guide the user through setup in this order:

No wallet detected
      │
      ▼
┌─────────────────────────────────────────┐
│ "Welcome to Thetanuts Options Trading!  │
│  Let's get you set up in 3 steps."      │
└─────────────────────────────────────────┘
      │
      ▼
┌─────────────────────────────────────────┐
│ Step 1: Create or import a wallet       │
│ • New user → wallet-create.js           │
│ • Have a seed? → wallet-import.js       │
│                                         │
│ WARN: Use a DEDICATED wallet.           │
│ Never reuse your primary wallet seed.   │
│ Custody stays LOCAL on your machine.    │
└─────────────────────────────────────────┘
      │
      ▼
┌─────────────────────────────────────────┐
│ Step 2: Verify & fund the wallet        │
│ • Run wallet-discover.js → show address │
│ • Run wallet-balance.js --chain         │
│   base-mainnet                          │
│ • If zero balance: explain funding      │
│   - Need ETH on Base for gas fees       │
│   - Need USDC for PUT trades            │
│   - Need WETH for CALL trades           │
│   - Bridge via bridge.base.org          │
└─────────────────────────────────────────┘
      │
      ▼
┌─────────────────────────────────────────┐
│ Step 3: Market overview & first trade   │
│ • Fetch prices: get-prices.ts           │
│ • WebSearch for market news             │
│ • Ask risk preference (LOW/MED/HIGH)    │
│ • Show first strategy recommendation    │
└─────────────────────────────────────────┘

Welcome Message (After Onboarding or Returning User)

Once wallet is configured, present:

Welcome to Thetanuts Options Trading!

Your wallet: 0x...
Chain: Base Mainnet
Balances: X ETH (gas) | Y USDC | Z WETH

What would you like to do?
• "Show me options"       → View available strikes and prices
• "Recommend a strategy"  → I'll check market news and suggest based on your risk preference
• "Check my positions"    → View open trades and P&L on expired ones
• "What's the market doing?" → Latest crypto news and how it affects options

New to options? Just say "teach me about options"

Security Reminders (shown during onboarding)

• DEDICATED WALLET: Use a dedicated wallet seed for this integration. Never reuse your primary wallet.
• LOCAL CUSTODY: Your seed phrase stays on your machine. The agent never sends it anywhere.
• TRANSACTIONS: Irreversible once broadcast. Always verify before sending.
• APPROVALS: Token approvals allow contracts to spend your tokens. Only approve trusted contracts.
• GAS: Ensure wallet has ETH on Base for gas fees.
• RFQ KEYS: Back up .thetanuts-keys/ directory — lost keys cannot decrypt past offers.

Thetanuts Options Trading

What is TNUT Agent?

TNUT Agent is your AI-powered assistant for trading crypto options on Thetanuts Finance.

What it does:

• Manages local wallets safely (custody stays on your machine)
• Accesses market intelligence (prices, orderbook, MM quotes)
• Inspects options and positions
• Prepares and executes trades through the Thetanuts protocol

Local-first design:

• Wallet custody stays LOCAL - your seed phrase never leaves your machine
• Signing stays LOCAL - transactions are signed on your device
• Backend provides intelligence, data, and routing support only

Capabilities:

Options 101

What are Options?

Options are contracts that give you the right (but not obligation) to buy or sell an asset at a specific price by a specific date.

PUT Options

A PUT gives the holder the right to SELL at the strike price.

Example:

• ETH is at $2000
• You BUY a $1900 PUT for $10 premium
• If ETH drops to $1800 → Your PUT is worth $100 (profit: $90)
• If ETH stays at $2000 → Your PUT expires worthless (loss: $10 premium)

CALL Options

A CALL gives the holder the right to BUY at the strike price.

Example:

• ETH is at $2000
• You BUY a $2100 CALL for $15 premium
• If ETH rises to $2300 → Your CALL is worth $200 (profit: $185)
• If ETH stays at $2000 → Your CALL expires worthless (loss: $15 premium)

Key Terms

Collateral Requirements by Option Type

IMPORTANT: Different option types require different collateral tokens:

• PUT options: Collateral in USDC. Formula: (collateral × 1e8) / strike
• CALL options: Collateral in WETH. Formula: collateral / 1e12 (1 WETH = 1 contract)

The build-rfq.ts script automatically selects the correct collateral based on option type.

Multi-Strike Option Structures

Beyond vanilla PUT and CALL options, Thetanuts supports advanced multi-strike structures for sophisticated trading strategies.

All Option Types

Strike Ordering Rules

CRITICAL: Strikes must be ordered correctly or the RFQ will fail.

Collateral Formulas by Structure (Human-Readable)

These are simplified formulas for quick estimation. For exact on-chain values, use previewFillOrder() or client.utils.calculateCollateral().

Key rule: Only INVERSE_CALL and PHYSICAL_CALL use WETH (base asset). Everything else uses USDC — including CALL_SPREAD, CALL_FLY, and CALL_CONDOR.

When to Use Each Structure

Risk-Categorized Strategy Recommendations

When a user asks for a strategy recommendation, FIRST ask about their risk preference, then match to appropriate strategies within that tier.

Step 1: Ask the User's Risk Level

Present these three options:

Step 2: Strategy Decision Tree by Risk Level

User wants a strategy recommendation
        │
        ▼
┌──────────────────────────────────┐
│ Step 1: Check market news        │
│ (WebSearch - see News section)   │
└──────────────────────────────────┘
        │
        ▼
┌──────────────────────────────────┐
│ Step 2: Ask risk level           │
│ LOW / MEDIUM / HIGH              │
└──────────────────────────────────┘
        │
        ├── LOW RISK
        │   │
        │   ├── Neutral view ──────► IRON_CONDOR (sell)
        │   │                        Collect premium both sides, profit if range-bound
        │   │                        Max loss: spread width - premium
        │   │                        Max gain: premium collected
        │   │                        Win prob: ~60-70% (price stays in range)
        │   │
        │   ├── Slightly bullish ──► PUT_FLY (buy) or Sell PUT via orderbook
        │   │                        Butterfly: tiny cost, big payout if pinned
        │   │                        Max loss: premium paid (very small)
        │   │                        Max gain: wing width - premium
        │   │
        │   ├── Slightly bearish ──► CALL_FLY (buy) or Sell CALL via orderbook
        │   │                        Same as above, opposite direction
        │   │
        │   └── Range-bound ───────► Butterfly (CALL_FLY / PUT_FLY)
        │                            Cheapest defined-risk play
        │                            Max loss: premium paid (~$3-5/contract)
        │                            Max gain: wing width (~$47-50/contract)
        │
        ├── MEDIUM RISK
        │   │
        │   ├── Bullish ──────────► CALL_SPREAD (buy)
        │   │                       Cheaper than vanilla call, capped both sides
        │   │                       Max loss: premium paid
        │   │                       Max gain: spread width - premium
        │   │                       Risk/Reward: typically 1:1 to 1:3
        │   │
        │   ├── Bearish ─────────► PUT_SPREAD (buy)
        │   │                       Cheaper than vanilla put, capped both sides
        │   │                       Max loss: premium paid
        │   │                       Max gain: spread width - premium
        │   │
        │   └── Range-bound ─────► CONDOR (CALL_CONDOR / PUT_CONDOR)
        │                          Wider profit range than butterfly
        │                          Max loss: wing width - premium collected
        │                          Max gain: premium collected (sell) or inner width (buy)
        │
        └── HIGH RISK
            │
            ├── Bullish ──────────► CALL (vanilla buy) / INVERSE_CALL
            │                       Unlimited profit potential if price rises
            │                       Max loss: full premium paid
            │                       Max gain: unlimited
            │
            ├── Bearish ─────────► PUT (vanilla buy)
            │                       Unlimited profit if price crashes
            │                       Max loss: full premium paid
            │                       Max gain: strike - premium (if price → 0)
            │
            └── Selling premium ──► Naked PUT sell or CALL sell
                                    Collect premium, very large risk if wrong
                                    Max loss: strike price (PUT) or unlimited (CALL)
                                    Max gain: premium collected

Risk Summary Card (Always Show When Recommending)

Every strategy recommendation MUST include this card:

┌────────────────────────────────────────────────┐
│ Strategy: [STRATEGY NAME]                      │
│ Risk Level: [LOW / MEDIUM / HIGH]              │
│ Direction: [BULLISH / BEARISH / NEUTRAL]       │
├────────────────────────────────────────────────┤
│ Max Loss:   $[amount] ([description])          │
│ Max Gain:   $[amount] ([description])          │
│ Break-even: $[price]                           │
│ Win Zone:   Price between $[low] and $[high]   │
│ Collateral: [amount] [token]                   │
└────────────────────────────────────────────────┘

Collateral Efficiency Comparison

Different structures give you different exposure per dollar of collateral:

Key Insight: For users with small collateral (under $100), LOW and MEDIUM risk strategies (spreads, butterflies) provide significantly more market exposure than HIGH risk vanilla options.

Strategy Selection Inputs

When recommending a strategy, consider these inputs in order of priority:

News & Events Integration

Before EVERY strategy recommendation, the agent MUST search for recent market-moving events. This ensures context-aware recommendations, not purely technical ones.

How to Fetch Market News

Use WebSearch with these queries (run at least 2):

If WebSearch returns a relevant article, use WebFetch to get deeper details.

Translating News to Strategy

News-Informed Recommendation Template

When presenting a strategy recommendation, always include market context:

Market Context (from recent news):
• [Event 1]: [brief description] → [bullish/bearish/neutral]
• [Event 2]: [brief description] → [bullish/bearish/neutral]

Overall market bias: [BULLISH / BEARISH / NEUTRAL / UNCERTAIN]

Based on [RISK LEVEL] risk preference and current market conditions:
[Strategy recommendation with Risk Summary Card]

Multi-Strike RFQ Examples

2-Strike Spread (CALL_SPREAD):

npx tsx scripts/build-rfq.ts \
  --underlying ETH \
  --type CALL \
  --strikes 2000,2100 \
  --expiry 1774684800 \
  --contracts 1 \
  --direction buy

Buy $2000 call, sell $2100 call. Max profit = $100 - premium. Max loss = premium.

3-Strike Butterfly (PUT_FLY):

npx tsx scripts/build-rfq.ts \
  --underlying ETH \
  --type PUT \
  --strikes 1900,1850,1800 \
  --expiry 1774684800 \
  --contracts 1 \
  --direction buy

Profit maximized if ETH settles at $1850. Low cost, high reward if pinned.

4-Strike Condor (IRON_CONDOR):

npx tsx scripts/build-rfq.ts \
  --underlying ETH \
  --type PUT \
  --strikes 1800,1900,2100,2200 \
  --expiry 1774684800 \
  --contracts 1 \
  --direction sell

Sell iron condor. Collect premium if ETH stays between $1900-$2100.

OptionBook vs RFQ (Factory)

The SDK supports two trading systems. Choose based on your use case:

Orderbook Trading

What it is: Fill existing orders posted by market makers - instant execution.

You → Fill existing order → Instant trade

Best for:

• Standard strikes/expiries with existing liquidity
• Quick trades when you see a good price
• Smaller to medium sizes

How it works:

1. Fetch orderbook: npx tsx scripts/fetch-orders.ts --type PUT
2. See available BIDs (buyers) and ASKs (sellers)
3. Fill the order you want: npx tsx scripts/fill-order.ts --order-index 0 --collateral 10 --seed "..." --execute

RFQ (Request for Quote)

What it is: Request a custom quote from market makers - they respond within 45 seconds.

You → Submit RFQ → MM sees it → MM sends encrypted offer → You settle (or accept early)

Best for:

• Custom strikes/expiries not on orderbook
• No existing liquidity at your terms
• Larger sizes (MMs may offer better pricing)
• Multi-leg structures (spreads, butterflies, condors)

How it works:

1. Build RFQ: npx tsx scripts/build-rfq.ts --underlying ETH --type PUT --strike 1900 --expiry <timestamp> --contracts 0.1 --direction buy
2. Send transaction with the returned to and data
3. Wait up to 45 seconds for MM response
4. Settle when offer received (or accept early - see Early Settlement)

Decision Matrix

Agent Decision Logic:

1. First, check orderbook for liquidity at your strike
2. If liquidity exists → recommend Orderbook
3. If no liquidity → recommend RFQ and explain why

Agent Decision Logic (Orderbook-First)

When a user wants to trade, the agent ALWAYS checks the orderbook first before recommending a trading method.

Decision Flowchart

User wants to trade
        │
        ▼
┌─────────────────────────┐
│  check-orderbook.ts     │
│  Check for liquidity    │
└─────────────────────────┘
        │
        ▼
┌─────────────────────────┐
│ Exact strike match?     │
└─────────────────────────┘
    │           │
   Yes          No
    │           │
    ▼           ▼
┌─────────┐  ┌─────────────────┐
│Orderbook│  │Nearby strikes?  │
│  Fill   │  │(within ±5%)     │
└─────────┘  └─────────────────┘
                 │         │
                Yes        No
                 │         │
                 ▼         ▼
           ┌─────────┐  ┌─────┐
           │Show     │  │ RFQ │
           │options  │  │     │
           └─────────┘  └─────┘

How to Use check-orderbook.ts

Before recommending orderbook or RFQ, run:

npx tsx scripts/check-orderbook.ts --underlying ETH --type PUT --strike 1900 --expiry 1774684800 --direction sell

The script returns:

• recommendation: "orderbook" or "rfq"
• reason: Explanation of why
• orderbookOrders: Matching orders with prices
• nearbyStrikes: Alternative strikes if exact not found
• nextStep: Command to run next

Example Scenarios

Scenario 1: Liquidity Found

User: "Sell a PUT at $1900"
Agent: Runs check-orderbook.ts
Result: {
  "recommendation": "orderbook",
  "reason": "Found orderbook liquidity at strike $1900. Best bid: $8.08. Available: 75 contracts."
}
Agent: "Found a buyer at $8.08/contract. Use orderbook for instant execution."

Scenario 2: No Exact Match, Nearby Available

User: "Sell a PUT at $1850"
Agent: Runs check-orderbook.ts
Result: {
  "recommendation": "rfq",
  "reason": "No orderbook liquidity at $1850. Nearby strikes: $1900 (+2.7%), $1925 (+4.1%)"
}
Agent: "No orders at $1850, but $1900 is available (+2.7%). Would you like that, or submit RFQ for $1850?"

Scenario 3: Partial Fill

User: "Sell 100 contracts at $1900"
Agent: Runs check-orderbook.ts --size 100
Result: {
  "recommendation": "orderbook",
  "partialFillAvailable": true,
  "partialSize": 75,
  "reason": "Found 75 contracts (you requested 100). Partial fill available."
}
Agent: "Only 75 contracts available on orderbook. Fill partial now, or use RFQ for full 100?"

Trading Workflows

Workflow 1: Fill an Orderbook Order (Instant)

Step 1: Check wallet
   └─> node scripts/wallet-discover.js

Step 2: Check balance
   └─> node scripts/wallet-balance.js --chain base-mainnet --tokens 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913

Step 3: Fetch orderbook
   └─> npx tsx scripts/fetch-orders.ts --type PUT

Step 4: Preview fill (see what you'll get)
   └─> npx tsx scripts/fill-order.ts --order-index 0 --collateral 10 --seed "..."

Step 5: Execute fill
   └─> npx tsx scripts/fill-order.ts --order-index 0 --collateral 10 --seed "..." --execute --wait

Workflow 2: Submit an RFQ (Custom Terms)

Step 1: Check wallet & balance (same as above)

Step 2: Get MM pricing to see market levels
   └─> npx tsx scripts/get-mm-pricing.ts ETH --type PUT

Step 3: Build RFQ with your terms
   └─> npx tsx scripts/build-rfq.ts --underlying ETH --type PUT --strike 1900 --expiry 1774684800 --contracts 0.1 --direction buy

Step 4: Approve tokens (if selling/first time)
   └─> npx tsx scripts/approve-token.ts --token 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913 --spender 0x1aDcD391CF15Fb699Ed29B1D394F4A64106886e5 --max --seed "..." --wait

Step 5: Send RFQ transaction
   └─> npx tsx scripts/send-transaction.ts --to <from step 3> --data <from step 3> --seed "..." --wait

Step 6: Wait for MM offer (up to 45 seconds)

Step 7: Verify RFQ fill
   └─> npx tsx scripts/check-rfq-fill.ts --address <wallet> --ticker <expected> --since <submission_timestamp>
   └─> If filled: Shows position details
   └─> If not filled: Shows suggestions for next steps

Workflow 3: Check Positions

npx tsx scripts/get-positions.ts 0xYourWalletAddress

Workflow 3b: Portfolio Performance & ROI Review

After trades expire, show the user their realized returns. This workflow uses existing scripts — no new tools needed.

How to Calculate ROI

Step 1: Fetch all positions
   └─> npx tsx scripts/get-positions.ts <wallet_address>

Step 2: Categorize by status
   - Settled/expired positions → REALIZED P&L
   - Active/open positions → UNREALIZED (show separately)

Step 3: For each expired position
   - Use the pnlUsd field from get-positions.ts (if available)
   - If pnlUsd is missing, calculate manually:
     └─> For PUT/CALL: npx tsx scripts/calculate-payout.ts --type <PUT|CALL> --strike <strike> --settlement <settlement_price> --contracts <amount> [--is-buyer]
     └─> For spreads/butterflies/condors: use client.option.calculatePayout(optionAddress, settlementPrice) on-chain
         (Note: calculate-payout.ts only supports vanilla PUT/CALL — multi-strike payouts need the SDK on-chain method)

Step 4: Aggregate and present using the template below

ROI Calculation Logic

For BUYERS (paid premium):

• Cost basis = premium paid (collateral committed)
• Return = pnlUsd or calculated payout
• ROI = (return / cost_basis) × 100%
• Expired worthless → ROI = -100%
• Example: Paid $10 premium, option settled for $25 payout → ROI = +150%

For SELLERS (received premium):

• Credit received = premium collected
• Settlement cost = pnlUsd (negative = you paid out)
• ROI = (net_pnl / collateral_locked) × 100%
• Expired worthless for buyer (good for seller) → ROI = +(premium / collateral) × 100%
• Example: Collected $8 premium on $100 collateral, expired OTM → ROI = +8%

Portfolio Performance Summary Template

Present results in this format:

Portfolio Performance Summary
══════════════════════════════════════════════════════════
Wallet: 0x...
Period: [earliest expired trade] to [latest expired trade]

REALIZED P&L (Expired Positions)
┌──────────────────────┬──────┬─────────┬─────────┐
│ Trade                │ Side │ P&L     │ ROI     │
├──────────────────────┼──────┼─────────┼─────────┤
│ ETH-28MAR-2000-P     │ SELL │ +$12.50 │ +6.25%  │
│ ETH-28MAR-2100-C     │ BUY  │ -$8.00  │ -100%   │
│ ETH-04APR-1900/1800  │ BUY  │ +$45.00 │ +150%   │
├──────────────────────┼──────┼─────────┼─────────┤
│ TOTAL                │      │ +$49.50 │         │
└──────────────────────┴──────┴─────────┴─────────┘

Win rate: 2/3 (66.7%)
Average P&L per trade: +$16.50
Best trade:  ETH PUT_SPREAD (+$45.00, +150%)
Worst trade: ETH CALL (-$8.00, expired worthless)

OPEN POSITIONS
┌──────────────────────┬──────┬───────────┬──────────┬────────┐
│ Trade                │ Side │ Contracts │ Expiry   │ Status │
├──────────────────────┼──────┼───────────┼──────────┼────────┤
│ ETH-18APR-2000-P     │ BUY  │ 0.5       │ Apr 18   │ Active │
└──────────────────────┴──────┴───────────┴──────────┴────────┘

When to Show Performance

• User asks "how are my trades doing?", "show my P&L", "what's my ROI?"
• User asks "check my positions" AND has expired positions → include ROI for expired ones
• Proactively mention performance when user has recently expired positions they haven't reviewed
• After any trade settles, offer to show the outcome: "Your ETH PUT just expired. Want to see how it did?"

Workflow 4: Early Settlement (Accept MM Offer Before Deadline)

After a market maker submits an encrypted offer to your RFQ, you can decrypt and accept it early — no need to wait for the full deadline.

Step 1: Submit RFQ (Workflow 2, Steps 1-5)

Step 2: MM sends encrypted offer (within 45 seconds)

Step 3: Decrypt the offer using your ECDH keypair
   └─> Uses client.rfqKeys.decryptOffer()

Step 4: Accept the offer early
   └─> Uses client.optionFactory.encodeSettleQuotationEarly(quotationId, offerAmount, nonce, offeror)
   └─> Send transaction with returned {to, data}

Real examples:

• RFQ 784 (PUT BUTTERFLY $1700/$1800/$1900): MM offered at 04:05:45 UTC, early settle at 04:07:09 UTC (3 min before deadline)
  • TX: 0x105f75cdfb64a3796100f6d667bc4f7fec3836d2b5aa5c43b66073a1b40964ee
• RFQ 785 (PUT CONDOR $1600/$1700/$1800/$1900): MM offered 0.003248 USDC, early settle at 04:15:00 UTC
  • TX: 0xa89fb6dbad43b430399bbdec878927185e602b7df9b5390f71d2d11c33e4d850

Workflow 5: Cancel RFQ & Claim Refund

If a user wants to cancel an active RFQ before it settles, or if no MM responded:

┌─────────────────────────────────────────────────────┐
│ RFQ submitted                                       │
│         │                                           │
│         ▼                                           │
│ ┌─────────────────────┐                             │
│ │ MM responded?       │                             │
│ └─────────────────────┘                             │
│     │           │                                   │
│    Yes          No                                  │
│     │           │                                   │
│     ▼           ▼                                   │
│ ┌─────────┐  ┌──────────────────────────────┐       │
│ │ Accept  │  │ RFQ expires with no offers   │       │
│ │ or      │  │ → No collateral was locked   │       │
│ │ Cancel? │  │ → Nothing to refund          │       │
│ └─────────┘  └──────────────────────────────┘       │
│   │     │                                           │
│  Accept Cancel                                      │
│   │     │                                           │
│   ▼     ▼                                           │
│ Settle  Cancel RFQ                                  │
│         └─> cancelQuotation(quotationId)            │
│         └─> Collateral returned to wallet           │
└─────────────────────────────────────────────────────┘

Cancel an Active RFQ

// Cancel before settlement — requester only
const { to, data } = client.optionFactory.encodeCancelQuotation(quotationId);
// Send transaction with {to, data} via send-transaction.ts

Cancel an Offer (Market Maker Side)

// MM cancels their offer before settlement
const { to, data } = client.optionFactory.encodeCancelOfferForQuotation(quotationId);

Check RFQ Status

// Query current state of an RFQ
const quotation = await client.optionFactory.getQuotation(quotationId);
// quotation status: 'active' | 'settled' | 'cancelled'
// quotation.isActive — whether RFQ is still open
// quotation.currentWinner — winning offeror's address

Refund Rules

IMPORTANT: Orderbook fills are INSTANT and IRREVERSIBLE. Only RFQs can be cancelled (before settlement).

Buy vs Sell Order: Collateral & Token Flow

No ETH is sent as transaction value for option trades. All collateral flows happen via ERC-20 token transfers (USDC, WETH). ETH is only used for gas fees.

When BUYING Options

Buyer Flow:
1. Premium is paid in collateral tokens (USDC for PUTs, WETH for CALLs)
2. Token approval needed: approve-token.ts (one-time per token/spender)
3. Collateral is transferred to the contract when the trade executes
4. If RFQ: collateralAmount = 0 in the request (factory holds it)
5. If Orderbook: collateral deducted at fill time

What you PAY:   Premium (in USDC or WETH) + gas (in ETH)
What you GET:   Option position (right to profit if price moves your way)
At expiry:      Payout if in-the-money, or nothing if out-of-the-money

When SELLING Options

Seller Flow:
1. Collateral is LOCKED to back the option (USDC for PUTs, WETH for CALLs)
2. Token approval needed: approve-token.ts (one-time per token/spender)
3. Collateral transferred to contract and held until expiry
4. Premium is received from buyer

What you PAY:   Collateral lock (USDC or WETH) + gas (in ETH)
What you GET:   Premium received upfront
At expiry:      Keep premium if OTM, or payout difference if ITM
                Remaining collateral returned after settlement

Token Requirements Summary

Always check balance BEFORE trading:

# Check USDC balance (for PUTs)
node scripts/wallet-balance.js --chain base-mainnet --tokens 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913

# Check WETH balance (for CALLs)
node scripts/wallet-balance.js --chain base-mainnet --tokens 0x4200000000000000000000000000000000000006

# Check ETH balance (for gas)
node scripts/wallet-balance.js --chain base-mainnet

Winner Selection & Payout by Option Type

The winner (winningOfferor) in an RFQ is always the market maker with the best offer price. Winner selection works the same way for ALL option types — PUT, CALL, spreads, butterflies, condors, iron condors, and physical options.

Winner Fields in SDK

The winner address appears in three places depending on context:

Payout Formulas by Option Type

SDK Payout Calculation Support

The SDK utility client.utils.calculatePayout() supports these types:

type PayoutType = 'call' | 'put' | 'call_spread' | 'put_spread';

For butterflies, condors, and iron condors: use client.option.calculatePayout(optionAddress, settlementPrice) which calculates on-chain for any option type, or use client.option.simulatePayout() for simulation.

// For vanilla and spreads — use utility function
const payout = client.utils.calculatePayout({
  type: 'call_spread',
  strikes: [2000n * 10n**8n, 2100n * 10n**8n],
  settlementPrice: 2050n * 10n**8n,
  numContracts: 1n * 10n**18n,
});

// For butterflies, condors, iron condors — use on-chain calculation
const payout = await client.option.calculatePayout(optionAddress, settlementPrice);

Settlement Events

When an option settles at expiry, these events are emitted:

Physical Settlement (PHYSICAL_CALL / PHYSICAL_PUT)

Physical options involve actual asset delivery instead of cash settlement:

The SDK has dedicated builders for physical options:

• client.optionFactory.buildPhysicalRFQ()
• client.optionFactory.buildPhysicalSpreadRFQ()
• client.optionFactory.buildPhysicalButterflyRFQ()
• client.optionFactory.buildPhysicalCondorRFQ()

Understanding Collateral vs Contracts

When viewing orders, availableAmount represents the maker's collateral budget, not the number of contracts. The actual number of purchasable contracts depends on the option type and collateral requirements.

Collateral Formulas by Option Type (On-Chain)

These are the on-chain formulas using raw decimal values. For practical use, prefer previewFillOrder() which handles all conversions automatically.

IMPORTANT: Only single-strike CALL options (INVERSE_CALL, PHYSICAL_CALL) use WETH collateral. ALL multi-strike structures use USDC, even CALL_SPREAD, CALL_FLY, and CALL_CONDOR.

Using previewFillOrder

Always use previewFillOrder() to see the actual contract count before filling:

const order = orders[0];

// Preview shows calculated max contracts based on collateral requirements
const preview = client.optionBook.previewFillOrder(order);
console.log(`Max contracts: ${preview.maxContracts}`);
console.log(`Collateral token: ${preview.collateralToken}`);
console.log(`Price per contract: ${preview.pricePerContract}`);

// Preview with specific premium amount
const preview10 = client.optionBook.previewFillOrder(order, 10_000000n); // 10 USDC premium
console.log(`Contracts for 10 USDC: ${preview10.numContracts}`);

Why This Matters

For a PUT option with a $95,000 strike:

• Maker provides: 10,000 USDC collateral
• Max contracts: 10,000 / 95,000 ≈ 0.105 contracts (not 10,000!)

The previewFillOrder() method handles these calculations automatically for all option types.

SDK Modules Reference

Referrer Fee Sharing

The SDK supports a referrer address for fee sharing on order fills. When a referrer is set, a portion of the trading fees is allocated to the referrer.

// Option 1: Set referrer at client initialization (applies to all fills)
const client = new ThetanutsClient({
  chainId: 8453,
  provider,
  signer,
  referrer: '0x92b8ac05b63472d1D84b32bDFBBf3e1887331567',
});

// All fillOrder calls will use this referrer automatically
await client.optionBook.fillOrder(order);

// Option 2: Pass referrer per fill call (overrides client default)
await client.optionBook.fillOrder(order, undefined, '0xYourReferrerAddress');

// Option 3: Use encode methods (for viem/wagmi/AA wallets)
const { to, data } = client.optionBook.encodeFillOrder(
  order,
  collateralAmount,
  '0x92b8ac05b63472d1D84b32bDFBBf3e1887331567'
);
const hash = await walletClient.sendTransaction({ to, data });

// Query referrer fee split
const feeBps = await client.optionBook.getReferrerFeeSplit('0x...');
console.log(`Referrer fee: ${feeBps} bps`);

// Query accumulated fees
const fees = await client.optionBook.getFees(usdcAddress, '0x...');
console.log(`Accumulated fees: ${fees}`);

If no referrer is provided, the zero address (0x000...) is used (no fee sharing).

RFQ Key Management

The SDK uses ECDH (Elliptic Curve Diffie-Hellman) key pairs for encrypted offers in the RFQ system. Keys are automatically persisted based on your environment:

Automatic Key Management

// Keys are automatically persisted - no configuration needed
const keyPair = await client.rfqKeys.getOrCreateKeyPair();
console.log('Public Key:', keyPair.compressedPublicKey);
// Keys are saved automatically and survive process restarts

Key Backup Warning

IMPORTANT: Back up your RFQ private keys! Keys are stored in .thetanuts-keys/ with secure permissions. If lost, you cannot decrypt offers made to your public key. There is no recovery mechanism.

Real-Time WebSocket Subscriptions

Subscribe to live updates for orders and prices:

const client = new ThetanutsClient({ chainId: 8453, provider });

// 1. Connect
await client.ws.connect();

// 2. Subscribe to order updates
const unsubOrders = client.ws.subscribeOrders((update) => {
  console.log(`Order ${update.event}:`, update);
});

// 3. Subscribe to price updates for ETH
const unsubPrices = client.ws.subscribePrices((update) => {
  console.log(`ETH price: $${update.price}`);
}, 'ETH');

// 4. Handle connection state changes
const unsubState = client.ws.onStateChange((state) => {
  console.log(`WebSocket state: ${state}`);
});

// 5. Disconnect when done
// unsubOrders(); unsubPrices(); unsubState();
// client.ws.disconnect();

The WebSocket module auto-reconnects by default (up to 10 attempts).

Error Handling

All SDK methods throw ThetanutsError with typed error codes:

import { isThetanutsError, OrderExpiredError, InsufficientAllowanceError } from '@thetanuts-finance/thetanuts-client';

try {
  await client.optionBook.fillOrder(order, 10_000000n);
} catch (error) {
  if (error instanceof OrderExpiredError) {
    console.log('Order expired, fetching fresh orders...');
    const freshOrders = await client.api.fetchOrders();
  } else if (error instanceof InsufficientAllowanceError) {
    console.log('Approving tokens first...');
    await client.erc20.ensureAllowance(usdcAddress, optionBookAddress, amount);
  } else if (isThetanutsError(error)) {
    switch (error.code) {
      case 'SLIPPAGE_EXCEEDED': console.log('Price moved too much'); break;
      case 'INSUFFICIENT_BALANCE': console.log('Not enough tokens'); break;
      case 'SIGNER_REQUIRED': console.log('Signer required'); break;
      case 'CONTRACT_REVERT': console.log('Contract call failed'); break;
    }
  }
}

Error Codes

SDK Configuration Options

interface ThetanutsClientConfig {
  chainId: 8453;                    // Required: Chain ID
  provider: Provider;               // Required: ethers.js provider
  signer?: Signer;                  // Optional: For transactions
  referrer?: string;                // Optional: Referrer address for fees
  apiBaseUrl?: string;              // Optional: Override API URL
  indexerApiUrl?: string;           // Optional: Override indexer URL
  pricingApiUrl?: string;           // Optional: Override pricing URL
  wsUrl?: string;                   // Optional: Override WebSocket URL
  env?: 'dev' | 'prod';             // Optional: Environment (default: prod)
  logger?: ThetanutsLogger;         // Optional: Custom logger
  keyStorageProvider?: StorageProvider; // Optional: Custom RFQ key storage
}

Custom Logger

const client = new ThetanutsClient({
  chainId: 8453,
  provider,
  logger: {
    debug: (msg, meta) => myLogger.debug(msg, meta),
    info: (msg, meta) => myLogger.info(msg, meta),
    warn: (msg, meta) => myLogger.warn(msg, meta),
    error: (msg, meta) => myLogger.error(msg, meta),
  },
});

Production Checklist

Before deploying to production, verify the following:

• RPC Provider: Use a reliable provider (Alchemy, Infura, QuikNode) instead of public https://mainnet.base.org which has strict rate limits
• Referrer Configuration: Set the referrer address to earn fee-sharing revenue on fills
• Error Logging: Pass a custom logger to capture errors in Sentry/Datadog
• Gas Buffer: SDK adds 20% gas buffer for Account Abstraction wallets (Coinbase Smart Wallet, Safe)
• Collateral Approval Flow: Always call client.erc20.ensureAllowance() before fillOrder()
• WebSocket Reconnection: Auto-reconnects up to 10 attempts by default
• Order Expiry Checks: Check order.expiry before filling to avoid wasted gas estimates
• RFQ Key Backup: Back up .thetanuts-keys/ directory — lost keys cannot decrypt past offers

Decimal Handling

The SDK provides utilities for safe decimal conversions:

// Convert to on-chain values
const usdc = client.utils.toBigInt('100.5', 6);   // 100500000n
const strike = client.utils.strikeToChain(1850);  // 185000000000n

// Convert from on-chain values
const display = client.utils.fromBigInt(100500000n, 6);  // '100.5'
const price = client.utils.strikeFromChain(185000000000n); // 1850

Example Conversations

"I just installed this, what do I do?"

Agent: [Triggers onboarding flow]

1. Runs wallet-discover.js → no wallet found
2. Enters onboarding mode
3. Guides through wallet creation
4. Shows balance and funding instructions
5. Fetches market news via WebSearch
6. Asks risk preference
7. Makes first strategy suggestion

Agent: Welcome to Thetanuts Options Trading! Let's get you set up.

I see you don't have a wallet configured yet.
Would you like to create a new wallet or import an existing one?

[After wallet creation...]

Your wallet: 0x7a3F...9e2B
Chain: Base Mainnet
Balance: 0 ETH | 0 USDC | 0 WETH

To start trading, you'll need to fund your wallet:
• ETH on Base for gas fees (bridge via bridge.base.org)
• USDC for PUT options | WETH for CALL options

Once funded, tell me:
1. What's your risk preference? (Low / Medium / High)
2. Are you bullish, bearish, or neutral on ETH?

I'll check the latest market news and recommend a strategy!

"What's happening in crypto that I should trade?"

Agent:

1. Uses WebSearch: "crypto market news today"
2. Uses WebSearch: "ethereum bitcoin price catalyst this week"
3. Summarizes findings and maps to strategy
4. Asks risk preference
5. Recommends with Risk Summary Card

Agent: Let me check the latest market news...

Market Context:
• Fed meeting tomorrow — markets expect rate hold, dovish tone → BULLISH
• Large BTC options expiry ($2.3B) on Friday → potential pin risk → NEUTRAL
• ETH Layer-2 TVL hitting ATH → positive sentiment → BULLISH

Overall bias: MODERATELY BULLISH with near-term volatility expected

What risk level are you comfortable with?

LOW:  Iron Condor — collect premium, profit if ETH stays in range
      Good for the expected range-bound action before Fed meeting
MEDIUM: Call Spread — capped upside play on the bullish bias
      Benefit from dovish Fed without unlimited downside
HIGH: Vanilla Call — unlimited upside if Fed sparks a rally
      Lose full premium if market sells the news

"Show me LOW risk strategies"

Agent:

1. Fetches current prices and MM pricing
2. Filters for LOW risk structures
3. Shows Risk Summary Cards for each

Agent: Here are the LOW RISK strategies available right now:

┌────────────────────────────────────────────────────────┐
│ Strategy: Iron Condor (ETH)                            │
│ Risk Level: LOW                                        │
│ Direction: NEUTRAL                                     │
├────────────────────────────────────────────────────────┤
│ Strikes: $1800 / $1900 / $2100 / $2200                │
│ Max Loss:   $85/contract                               │
│ Max Gain:   $15/contract (premium collected)           │
│ Win Zone:   ETH stays between $1900-$2100              │
│ Current ETH: $2,050 (right in the sweet spot!)         │
│ Collateral: ~$100 USDC per contract                    │
└────────────────────────────────────────────────────────┘

┌────────────────────────────────────────────────────────┐
│ Strategy: Put Butterfly (ETH)                          │
│ Risk Level: LOW                                        │
│ Direction: NEUTRAL (target $2,000)                     │
├────────────────────────────────────────────────────────┤
│ Strikes: $1950 / $2000 / $2050                         │
│ Max Loss:   ~$3/contract (premium paid)                │
│ Max Gain:   ~$47/contract (if ETH settles at $2,000)   │
│ Win Zone:   ETH settles near $2,000                    │
│ Collateral: ~$50 USDC per contract                     │
└────────────────────────────────────────────────────────┘

Which strategy interests you? Or want to see MEDIUM or HIGH risk options?

"How did my trades do?"

Agent:

1. Runs get-positions.ts with user's wallet address
2. Separates expired from active positions
3. Computes ROI for each expired position
4. Presents portfolio performance summary

Agent: Let me check your trading performance...

Portfolio Performance Summary
══════════════════════════════════════════════════════════
Wallet: 0x7a3F...9e2B
Period: Mar 14, 2026 to Apr 11, 2026

REALIZED P&L (Expired Positions)
┌───────────────────────┬──────┬──────────┬─────────┐
│ Trade                 │ Side │ P&L      │ ROI     │
├───────────────────────┼──────┼──────────┼─────────┤
│ ETH-28MAR-2000-P      │ SELL │ +$12.50  │ +6.25%  │
│ ETH-28MAR-2100-C      │ BUY  │ -$8.00   │ -100%   │
│ ETH-04APR-1900/1800-PS│ BUY  │ +$45.00  │ +150%   │
├───────────────────────┼──────┼──────────┼─────────┤
│ TOTAL                 │      │ +$49.50  │         │
└───────────────────────┴──────┴──────────┴─────────┘

Win rate: 2/3 (66.7%)
Average P&L per trade: +$16.50
Best trade:  ETH PUT_SPREAD (+$45.00, +150% ROI)
Worst trade: ETH CALL (-$8.00, expired worthless)

Insight: Your MEDIUM risk spread trades outperformed the HIGH risk
vanilla call. Consider focusing on spreads for better risk-adjusted returns.

OPEN POSITIONS
┌───────────────────────┬──────┬───────────┬──────────┐
│ Trade                 │ Side │ Contracts │ Expiry   │
├───────────────────────┼──────┼───────────┼──────────┤
│ ETH-18APR-2000-P      │ BUY  │ 0.5       │ Apr 18   │
└───────────────────────┴──────┴───────────┴──────────┘

Want me to analyze what your open position might settle at?

"Show me ETH puts"

User: Show me ETH puts

Agent:

1. Fetches MM pricing: npx tsx scripts/get-mm-pricing.ts ETH --type PUT
2. Shows available strikes, expiries, bid/ask prices
3. Asks which one interests you

"I want to sell a put at $1900 with 10 USDC"

Agent's Decision Process:

1. Check orderbook for $1900 PUT bids
2. Found liquidity? → "Use Orderbook - there's a BID at $1900"
3. No liquidity? → "Use RFQ - I'll request a quote for you"

If Orderbook:

Agent: There's a BID at $1900. Let me fill it for you.
1. Previews: "You'll sell ~0.005 contracts, receive ~$X premium"
2. Asks: "Proceed?"
3. Executes fill

If RFQ:

Agent: No orderbook liquidity at $1900. I'll submit an RFQ.
1. Builds RFQ for ETH-1900-PUT sell
2. Approves USDC if needed
3. Submits RFQ
4. "RFQ submitted. MMs have 45 seconds to respond."

"Check my positions"

Agent:

1. Gets wallet address from discover
2. Runs: npx tsx scripts/get-positions.ts 0xYourAddress
3. Shows open positions with current P&L

"I have $50 and I'm bearish on ETH, recommend a strategy"

Agent's Decision Process:

1. WebSearch for market news → confirms bearish sentiment
2. User has small collateral ($50) → Consider spreads for efficiency
3. User is bearish → PUT direction
4. Ask risk level → user says "medium"
5. Recommendation: PUT_SPREAD (MEDIUM RISK)

Agent:

Market Context:
• ETH down 3% on regulatory concerns → BEARISH
• Overall market sentiment weak → confirms bearish view

With $50 and a bearish view at MEDIUM risk, I recommend a PUT SPREAD:

┌────────────────────────────────────────────────────────┐
│ Strategy: PUT_SPREAD        Risk: MEDIUM               │
│ Direction: BEARISH                                     │
├────────────────────────────────────────────────────────┤
│ Max Loss:   $50 (your full collateral)                 │
│ Max Gain:   $50 (spread width - premium)               │
│ Break-even: ~$1892                                     │
│ Win Zone:   ETH below $1900 at expiry                  │
│ Collateral: $50 USDC → 0.5 contracts (20x more than   │
│             vanilla PUT which gives only 0.026)        │
└────────────────────────────────────────────────────────┘

Available strikes for tomorrow's expiry:
• PUT_SPREAD $1950/$1850 - Bid: $12/contract
• PUT_SPREAD $1900/$1800 - Bid: $8/contract

Which spread would you like?

Want HIGH risk instead? Vanilla PUT gives unlimited profit but only 0.026 contracts.
Want LOW risk instead? Put Butterfly costs ~$3 with $47 max gain.

"I think ETH will stay around $2000 this week"

Agent's Decision Process:

1. WebSearch for news → no major catalysts, range-bound expected
2. User expects neutral/range-bound market
3. Ask risk level → user says "low"
4. Check collateral: $30 → butterfly is most efficient
5. Recommendation: CALL_FLY (LOW RISK)

Agent:

Market Context:
• No major catalysts this week → NEUTRAL
• Options expiry Friday with large OI at $2000 → pin risk → NEUTRAL

With $30 and a neutral view, here's my LOW RISK recommendation:

┌────────────────────────────────────────────────────────┐
│ Strategy: CALL_FLY (Butterfly)   Risk: LOW             │
│ Direction: NEUTRAL (target $2,050)                     │
├────────────────────────────────────────────────────────┤
│ Strikes: 2000 / 2050 / 2100                            │
│ Max Loss:   ~$5/contract (premium paid)                │
│ Max Gain:   ~$45/contract (at exactly $2050)           │
│ Break-even: $2005 - $2095                              │
│ Collateral: $30 → ~6 contracts                         │
└────────────────────────────────────────────────────────┘

Why butterfly (LOW risk)?
• Cheapest way to play range-bound (more contracts per dollar)
• Max loss is tiny (~$5/contract)
• If ETH pins near $2050, each contract pays ~$45

Want MEDIUM risk? A Condor ($100+) gives a wider profit zone.

Build this RFQ? (y/n)

"I want to sell premium and collect income"

Agent's Decision Process:

1. WebSearch for news → range-bound market expected
2. User wants to sell premium → Income strategy
3. Risk level: LOW (income/premium selling = defined risk)
4. If neutral → IRON_CONDOR

Agent:

Market Context:
• Market consolidating after rally → NEUTRAL
• No major events this week → supports range-bound play

For premium selling, I recommend an IRON CONDOR (LOW RISK):

┌────────────────────────────────────────────────────────┐
│ Strategy: IRON_CONDOR (sell)     Risk: LOW             │
│ Direction: NEUTRAL (profit if range-bound)             │
├────────────────────────────────────────────────────────┤
│ Strikes: $1800 / $1900 / $2100 / $2200                │
│ Max Loss:   $85/contract (if ETH exits range)          │
│ Max Gain:   $15/contract (premium collected)           │
│ Win Zone:   ETH stays between $1900-$2100              │
│ Current ETH: $2,000 (right in the middle!)             │
│ Collateral: ~$100 USDC per contract                    │
└────────────────────────────────────────────────────────┘

This strategy profits ~60-70% of the time in range-bound markets.
You collect $15 premium and risk $85 if wrong.

Is ETH likely to stay between $1900-$2100? Let me build this RFQ.

Output Formatting & UX Guidelines

Progressive Disclosure

Tailor detail level to the user's experience:

Formatting Standards

For strategy recommendations — Always include the Risk Summary Card (see "Risk-Categorized Strategy Recommendations" section for the full template with all fields: Strategy, Risk Level, Direction, Max Loss, Max Gain, Break-even, Win Zone, Collateral).

For price/market data — Use compact tables:

ETH Options (PUT) — Current Price: $2,050
┌────────────┬─────────┬────────┬────────┐
│ Strike     │ Expiry  │ Bid    │ Ask    │
├────────────┼─────────┼────────┼────────┤
│ $1,900     │ Apr 25  │ $8.50  │ $12.00 │
│ $2,000     │ Apr 25  │ $25.00 │ $30.00 │
└────────────┴─────────┴────────┴────────┘

For trade confirmations — Clear summary before execution:

Trade Summary
═════════════
Action: Buy PUT_SPREAD
Risk Level: MEDIUM
Strikes: $1900 / $1800
Expiry: Apr 25, 2026
Contracts: 1.0
Cost: ~$15 USDC
Max Loss: $15 | Max Gain: $85

Confirm? (yes/no)

For errors — Never show raw JSON to the user:

• Translate error codes to plain English
• Always suggest a corrective action
• Example: Instead of {"error": "INSUFFICIENT_ALLOWANCE"} → "You need to approve USDC spending first. Want me to do that?"

MCP Tool Usage for Enhanced UX

Context7 for SDK Documentation

When a user asks a technical question about the Thetanuts SDK:

1. Use context7 resolve-library-id to find thetanuts-finance/thetanuts-client
2. Use context7 query-docs with the specific topic (e.g., "fillOrder", "WebSocket", "error handling")
3. Provide accurate, up-to-date answers from official documentation

Commands Reference

Wallet Commands

Market Data Commands

Trading Commands

Position Commands

Contract Addresses (Base Mainnet)

Core Contracts

Tokens

Option Implementations

Price Feeds (Chainlink)

Updates

Updates are checked automatically at the start of every conversation (see First-Time Setup). The agent runs update.sh silently and only notifies the user if a new version is installed.

To manually check or force dependency updates:

bash scripts/update.sh

Optional flags:

• REFRESH_WDK_DEPS=1 - Refresh dependencies from lockfile
• UPGRADE_WDK_DEPS=1 - Upgrade dependency versions
• RESTART_WDK_RUNTIME=1 - Best-effort restart of WDK runtime

Updates NEVER modify wallet secrets (.env, WDK_SEED).

Ticker Format

Options use: {UNDERLYING}-{EXPIRY}-{STRIKE}-{TYPE}

Examples:

• ETH-28MAR26-2500-P = ETH Put, $2500 strike, March 28 2026 expiry
• BTC-28MAR26-95000-C = BTC Call, $95000 strike, March 28 2026 expiry

Network

• Chain: Base Mainnet (Chain ID 8453)
• Collateral: USDC (6 decimals), WETH (18 decimals), cbBTC (8 decimals)
• Strikes: 8 decimals internally
• Expiry: 8:00 UTC on expiry date