Execute perpetual trades on Ostium, Aster, and Avantis via Maxxit's Lazy Trading API, and trade Indian stocks through Zerodha Kite. Includes programmatic endpoints for opening/closing positions, managing risk, fetching market data, researching Indian equities, copy-trading other OpenClaw agents, and a trustless Alpha Marketplace for buying/selling ZK-verified trading signals (Arbitrum Sepolia).
Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
Execute perpetual trades on Ostium, Aster, and Avantis via Maxxit's Lazy Trading API, and trade Indian stocks through Zerodha Kite. Includes programmatic endpoints for opening/closing positions, managing risk, fetching market data, researching Indian equities, copy-trading other OpenClaw agents, and a trustless Alpha Marketplace for buying/selling ZK-verified trading signals (Arbitrum Sepolia).
Execute perpetual futures trades on Ostium, Aster DEX, and Avantis DEX through Maxxit's Lazy Trading API. This skill enables automated trading through programmatic endpoints for opening/closing positions and managing risk.
Built-in Strategy Scripts
The skill includes standalone Python strategy scripts. Use them when the user wants the agent to run a predefined trading system instead of manually specifying each trade.
ema-strategy.py
Trend-following EMA crossover on Binance klines using close prices.
rsi-bollinger-strategy.py
Mean-reversion system that waits for price to pierce a Bollinger Band and re-enter with RSI confirmation.
donchian-adx-strategy.py
Breakout system that trades Donchian channel breaks only when ADX confirms a strong trend regime.
mean-reversion-strategy.py - RSI + Bollinger Band mean-reversion strategy. A technical approach using price exhaustion points optimized for high-frequency scalping in sideways or boring markets.
breakout-strategy.py - Volatility breakout strategy with ATR filter. Enters trades when price breaks out of a standard deviation channel while ATR confirms increasing volatility and momentum.
vwap-strategy.py - VWAP crossover institutional momentum strategy. Uses volume-weighted average price and EMA to confirm institutional trend alignment and confirm trade strength with volume.
All scripts:
read Binance kline data directly from https://api.binance.com/api/v3/klines
use MAXXIT_API_URL and MAXXIT_API_KEY
execute through Maxxit programmatic trading endpoints
maintain per-symbol, per-venue state in the OpenClaw workspace
Always ask venue first if unclear: "Do you want to trade on Ostium, Aster, or Avantis?"
Always state the active venue explicitly in your response (e.g., "Using Ostium..." or "Using Aster..." or "Using Avantis...").
Do not mix venue suggestions:
If user is trading on Ostium, only suggest Ostium endpoints/actions.
If user is trading on Aster, only suggest Aster endpoints/actions.
If user is trading on Avantis, only suggest Avantis endpoints/actions.
Do not ask network clarification:
Ostium defaults to mainnet, but if the user explicitly asks for Ostium testnet / Arbitrum Sepolia, honor that and pass isTestnet: true on Ostium endpoints.
Aster is testnet-only in this setup.
Avantis is mainnet-only (Base chain) in this setup.
Therefore do not ask "mainnet or testnet?" unless the user explicitly requests Ostium testnet.
If user switches venue mid-conversation, confirm the switch and then continue with only that venue's flow.
⚠️ CRITICAL: API Parameter Rules (Read Before Calling ANY Endpoint)
NEVER assume, guess, or hallucinate values for API request parameters. Every required parameter must come from either a prior API response or explicit user input. If you don't have a required value, you MUST fetch it from the appropriate dependency endpoint first.
Parameter Dependency Graph
The following shows where each required parameter comes from. Always resolve dependencies before calling an endpoint.
Always call /user-details first to get user_wallet (used as userAddress/address). Cache it for the session — it doesn't change.
Treat /user-details as identity-first. It always returns user_wallet for a valid API key, even if no lazy-trading agent exists yet.
/user-details is sparse. It omits fields that are empty, null, or false. Missing fields mean “not applicable” and should not be treated as an error or missing configuration by themselves.
Only use ostium_agent_address when the venue needs an agent. Ostium and Avantis require it. Zerodha does not. Aster only needs user_wallet, but aster_configured must be present and true.
Never hardcode or guess wallet addresses. They are unique per user and must come from /user-details.
For opening a position: Fetch current market context first (via /api/lazy-trading/research, /api/lazy-trading/indian-stocks, /market-data, or /price as appropriate), present it to the user, get explicit confirmation plus trade parameters (collateral, leverage, side, TP, SL), then execute.
Market format rule (Ostium):/symbols returns pairs like ETH/USD, but /open-position expects market as base token only (e.g. ETH). Convert by taking the base token before /.
For setting TP/SL after opening: Use the actualTradeIndex from the /open-position response. If you don't have it (e.g., position was opened earlier), call /positions to get tradeIndex, pairIndex, and entryPrice.
For closing a position: You need the tradeIndex — always call /positions first to look up the correct one for the user's specified market/position.
Ask the user for trade parameters — never assume collateral amount, leverage, TP%, or SL%. Present defaults but let the user confirm or override.
Validate the market exists by calling /symbols before trading if you're unsure whether a token is available on Ostium.
For Alpha consumer flow: Follow the exact order: /alpha/agents → /alpha/listings → /alpha/purchase (402) → /alpha/pay → /alpha/purchase (with X-Payment) → /alpha/verify → /user-details → /alpha/execute. Never skip steps. For /alpha/verify, pass the content object exactly as received from purchase — do not modify keys or values.
Pre-Flight Checklist (Run Mentally Before Every API Call)
✅ Do I have the user's wallet address? → If not, call /user-details
✅ Does this flow require an agent address? → If yes, call /user-details and verify ostium_agent_address is present
✅ Does this endpoint need a tradeIndex? → If not in hand, call /positions
✅ Does this endpoint need entryPrice/pairIndex? → If not in hand, call /positions
✅ Did I ask the user for all trade parameters? → collateral, leverage, side, TP%, SL%
✅ Is the market/symbol valid? → If unsure, call /symbols to verify
✅ (Alpha) Do I have commitment? → If not, call /alpha/agents
✅ (Alpha) Do I have listingId? → If not, call /alpha/listings
✅ (Alpha) For /verify: Am I passing content exactly as received? → No modifications
✅ (Alpha) For /execute: Do I have agentAddress + userAddress? → Call /user-details
Authentication
All requests require an API key with prefix lt_. Pass it via:
Header: X-API-KEY: lt_your_api_key
Or: Authorization: Bearer lt_your_api_key
Market Research Workflow
When the user asks for market research, use the Maxxit market research endpoint instead of writing the research from scratch.
Endpoint:
POST /api/lazy-trading/research
POST /api/lazy-trading/indian-stocks for Indian equities research queries
Rules:
Construct the content prompt from the user's ask.
Preserve the user's asset, timeframe, strategy, and risk focus.
If the user is vague, build a best-effort trading research query from the context they gave instead of inventing a different objective.
Prefer prompts that ask for market structure, trend, momentum, support/resistance, catalysts, and trading risks when relevant.
Set deepResearch to true when the user asks for deep research, a comprehensive comparison, a detailed diligence-style breakdown, or explicitly wants more thorough research.
Set deepResearch to false for standard market summaries, quick trade briefs, or normal tactical research requests.
For POST /api/lazy-trading/indian-stocks, OpenClaw must decide the request options from the user's query and should not ask the user to choose chat_model, response_length, or thinking_level.
Always send question plus the inferred request options:
chat_model: analytical or strategic
response_length: short, medium, or long
thinking_level: only include this when chat_model is strategic
Use chat_model: "strategic" when the user wants a trading plan, swing-trade setups, long/short ideas, entry zones, stop loss, target levels, timing for this week, or tactical positioning. In strategic mode, include thinking_level automatically:
use balanced by default
use low for quick, lightweight tactical asks
use deep only when the user explicitly asks for a more thoughtful or more detailed strategy answer
Use chat_model: "analytical" for screens, rankings, fundamentals, valuation, sector comparisons, capex-cycle beneficiaries, earnings quality, balance sheet analysis, and diligence-style research. Do not send thinking_level in analytical mode.
Infer response_length from the user’s ask:
use short for quick answers, concise trade ideas, and direct setup requests
use medium by default for normal research requests
use long for ranked lists, detailed comparisons, deep dives, or multi-factor explanations
If the user’s wording contains both analytical and tactical elements, prioritize the main deliverable. If the answer must provide actionable trade setups, choose strategic; if the answer is mainly screening, ranking, valuation, or fundamental comparison, choose analytical.
Summarize the response and format it for readability.
Prompt construction examples:
User: "Research BTC for a swing long."
Query: Analyze BTC for a swing-long setup. Cover market structure, momentum, key support/resistance, likely catalysts, invalidation levels, and major trading risks.
User: "Give me market research on ETH for today."
Query: Summarize ETH market structure for today, including trend, momentum, key support/resistance, important catalysts, and trading risks for intraday positioning.
User: "Research SOL before I short it."
Query: Analyze SOL for a potential short setup. Cover current market structure, weakness signals, resistance levels, downside levels to watch, catalysts, and key squeeze/invalidation risks.
Example call:
curl -L -X POST "${MAXXIT_API_URL}/api/lazy-trading/research" \
-H "X-API-KEY: ${MAXXIT_API_KEY}" \
-H 'Content-Type: application/json' \
-d '{
"content": "Analyze BTC for a swing-long setup. Cover market structure, momentum, key support/resistance, likely catalysts, invalidation levels, and major trading risks.",
"deepResearch": false
}'
Indian stocks example:
curl -L -X POST "${MAXXIT_API_URL}/api/lazy-trading/indian-stocks" \
-H "X-API-KEY: ${MAXXIT_API_KEY}" \
-H 'Content-Type: application/json' \
-d '{
"question": "Screen for Indian IT stocks with strong profit growth and low debt.",
"chat_model": "analytical",
"response_length": "medium"
}'
Indian stocks tactical example:
curl -L -X POST "${MAXXIT_API_URL}/api/lazy-trading/indian-stocks" \
-H "X-API-KEY: ${MAXXIT_API_KEY}" \
-H 'Content-Type: application/json' \
-d '{
"question": "Which Indian banking stocks look strongest for a swing trade this week? Give long ideas only, with entry zone, stop loss, target range, and the reasoning behind each setup.",
"chat_model": "strategic",
"response_length": "short",
"thinking_level": "balanced"
}'
When discussing Indian equities, NSE/BSE orders, holdings, targets, stop losses, or portfolio values:
use Indian rupees as the default unit
prefer ₹ in user-facing responses (for example, ₹2,450, ₹1.2 lakh)
Retrieve USDC and ETH balance for the user's Ostium wallet address.
⚠️ Dependency: The address field is the user's Ostium wallet address (user_wallet). You MUST fetch it from /user-details first — do NOT hardcode or assume any address.
Get all open positions for the user's Ostium trading account. This endpoint is critical — it returns tradeIndex, pairIndex, and entryPrice which are required for closing positions and setting TP/SL.
⚠️ Dependency: The address field must come from /user-details → user_wallet. NEVER guess it.
tradeIndex — needed for /close-position, /set-take-profit, /set-stop-loss
pairIndex — needed for /set-take-profit, /set-stop-loss
entryPrice — needed for /set-take-profit, /set-stop-loss
side — needed for /set-take-profit, /set-stop-loss
### Get Position History
Get trading history for a wallet.
- `venue: "OSTIUM"` (default): uses Ostium history.
- `venue: "AVANTIS"`: returns normalized closed-trade history from Avantis `v2/history/portfolio/history`.
**Note:** The user's Ostium wallet address can be fetched from the `/api/lazy-trading/programmatic/user-details` endpoint (see Get Account Balance section above).
```bash
curl -L -X POST "${MAXXIT_API_URL}/api/lazy-trading/programmatic/history" \
-H "X-API-KEY: ${MAXXIT_API_KEY}" \
-H "Content-Type: application/json" \
-d '{"venue":"OSTIUM","address":"0x...","count":50}'
Request Body:
{"venue":"OSTIUM",// Optional: "OSTIUM" (default) or "AVANTIS""address":"0x...",// Required for OSTIUM; also accepted for AVANTIS as alias of userAddress"count":50// Number of recent orders to retrieve (default: 50)}
⚠️ Dependencies — ALL must be resolved BEFORE calling this endpoint:
agentAddress → from /user-details → ostium_agent_address (NEVER guess)
userAddress → from /user-details → user_wallet (NEVER guess)
market → validate via /symbols endpoint if unsure the token exists
If /symbols returns ETH/USD, pass market: "ETH" to /open-position (not ETH/USD)
side, collateral, leverage → ASK the user explicitly, do not assume
📊 Recommended Pre-Trade Flow:
Call /api/lazy-trading/research for crypto trade research, or /market-data / /price for current market conditions
Present the market context to the user (price, structure, momentum, volatility when available)
Ask the user: "Do you want to proceed? Specify: collateral (USDC), leverage, long/short"
Only after user confirms → call /open-position
🔐 Verification Note: Every trade is analyzed by EigenAI for alignment with market conditions. Users can verify the cryptographic signatures and reasoning for all their trades at maxxit.ai/openclaw.
🔑 SAVE the response — actualTradeIndex and entryPrice are needed for setting TP/SL later.
{"agentAddress":"0x...",// REQUIRED — from /user-details → ostium_agent_address. NEVER guess."userAddress":"0x...",// REQUIRED — from /user-details → user_wallet. NEVER guess."market":"BTC",// REQUIRED — Base token only for Ostium (e.g. "ETH", not "ETH/USD"). Validate via /symbols if unsure."side":"long",// REQUIRED — "long" or "short". ASK the user."collateral":100,// REQUIRED — Collateral in USDC. ASK the user."leverage":10,// Optional (default: 10). ASK the user."deploymentId":"uuid...",// Optional — associated deployment ID"signalId":"uuid...",// Optional — associated signal ID"isTestnet":false// Optional. Set true only when user explicitly asks for Ostium testnet / Arbitrum Sepolia.}
Response (IMPORTANT — save these values):
{"success":true,"orderId":"order_123","tradeId":"trade_abc","transactionHash":"0x...","txHash":"0x...","status":"OPEN","message":"Position opened successfully","actualTradeIndex":2,// ← SAVE THIS — needed for /set-take-profit and /set-stop-loss"entryPrice":95000.0,// ← SAVE THIS — needed for /set-take-profit and /set-stop-loss"reasoning":"Market sentiment is bullish...",// EigenAI trade alignment analysis"llmSignature":"0x..."// Cryptographic signature for auditability}
Close Position
Close an existing perpetual futures position on Ostium.
⚠️ Dependencies — resolve BEFORE calling this endpoint:
agentAddress → from /user-details → ostium_agent_address
userAddress → from /user-details → user_wallet
tradeIndex → call /positions first to find the position you want to close, then use its tradeIndex
NEVER guess the tradeIndex or tradeId. Always fetch from /positions endpoint.
{"agentAddress":"0x...",// REQUIRED — from /user-details → ostium_agent_address. NEVER guess."userAddress":"0x...",// REQUIRED — from /user-details → user_wallet. NEVER guess."market":"BTC",// REQUIRED — Token symbol"tradeId":"12345",// Optional — from /positions → tradeId"actualTradeIndex":2,// Highly recommended — from /positions → tradeIndex. NEVER guess."isTestnet":false// Optional. Set true only when user explicitly asks for Ostium testnet / Arbitrum Sepolia.}
Set or update take-profit level for an existing position on Ostium.
⚠️ Dependencies — you need ALL of these before calling:
agentAddress → from /user-details → ostium_agent_address
userAddress → from /user-details → user_wallet
tradeIndex → from /open-position response → actualTradeIndex, OR from /positions → tradeIndex
entryPrice → from /open-position response → entryPrice, OR from /positions → entryPrice
pairIndex → from /positions → pairIndex, OR from /symbols → symbol id
takeProfitPercent → ASK the user (default: 0.30 = 30%)
side → from /positions → side ("long" or "short")
If you just opened a position: Use actualTradeIndex and entryPrice from the /open-position response.
If the position was opened earlier: Call /positions to fetch tradeIndex, entryPrice, pairIndex, and side.
{"agentAddress":"0x...",// REQUIRED — from /user-details. NEVER guess."userAddress":"0x...",// REQUIRED — from /user-details. NEVER guess."market":"BTC",// REQUIRED — Token symbol"tradeIndex":2,// REQUIRED — from /open-position or /positions. NEVER guess."takeProfitPercent":0.30,// Optional (default: 0.30 = 30%). ASK the user."entryPrice":90000,// REQUIRED — from /open-position or /positions. NEVER guess."pairIndex":0,// REQUIRED — from /positions or /symbols. NEVER guess."side":"long",// Optional (default: "long") — from /positions."isTestnet":false// Optional. Set true only when user explicitly asks for Ostium testnet / Arbitrum Sepolia.}
Response:
{"success":true,"message":"Take profit set successfully","tpPrice":117000.0}
Set Stop Loss
Set or update stop-loss level for an existing position on Ostium.
⚠️ Dependencies — identical to Set Take Profit. You need ALL of these before calling:
agentAddress → from /user-details → ostium_agent_address
userAddress → from /user-details → user_wallet
tradeIndex → from /open-position response → actualTradeIndex, OR from /positions → tradeIndex
entryPrice → from /open-position response → entryPrice, OR from /positions → entryPrice
pairIndex → from /positions → pairIndex, OR from /symbols → symbol id
stopLossPercent → ASK the user (default: 0.10 = 10%)
side → from /positions → side ("long" or "short")
If you just opened a position: Use actualTradeIndex and entryPrice from the /open-position response.
If the position was opened earlier: Call /positions to fetch tradeIndex, entryPrice, pairIndex, and side.
# Same dependency resolution as Set Take Profit (see above for full example)# Step 1: Get addresses from /user-details# Step 2: Get position details from /positions# Step 3: Set stop loss with user-specified stopLossPercent
curl -L -X POST "${MAXXIT_API_URL}/api/lazy-trading/programmatic/set-stop-loss" \
-H "X-API-KEY: ${MAXXIT_API_KEY}" \
-H "Content-Type: application/json" \
-d '{
"agentAddress": "0x...",
"userAddress": "0x...",
"market": "BTC",
"tradeIndex": 2,
"stopLossPercent": 0.10,
"entryPrice": 90000,
"pairIndex": 0,
"side": "long"
}'
Request Body:
{"agentAddress":"0x...",// REQUIRED — from /user-details. NEVER guess."userAddress":"0x...",// REQUIRED — from /user-details. NEVER guess."market":"BTC",// REQUIRED — Token symbol"tradeIndex":2,// REQUIRED — from /open-position or /positions. NEVER guess."stopLossPercent":0.10,// Optional (default: 0.10 = 10%). ASK the user."entryPrice":90000,// REQUIRED — from /open-position or /positions. NEVER guess."pairIndex":0,// REQUIRED — from /positions or /symbols. NEVER guess."side":"long",// Optional (default: "long") — from /positions."isTestnet":false// Optional. Set true only when user explicitly asks for Ostium testnet / Arbitrum Sepolia.}
Response:
{"success":true,"message":"Stop loss set successfully","slPrice":81000.0,"liquidationPrice":85500.0,"adjusted":false}
Get All Market Data
Retrieve the complete market snapshot from Ostium, including all symbols and their current metrics. This is useful for market-wide scanning or analysis in a single request.
curl -L -X GET "${MAXXIT_API_URL}/api/lazy-trading/programmatic/market-data" \
-H "X-API-KEY: ${MAXXIT_API_KEY}"
Discover other OpenClaw Traders and top-performing traders to potentially copy-trade. This is the first step in the copy-trading workflow — the returned wallet addresses are used as the address parameter in the /copy-trader-trades endpoint.
⚠️ Dependency Chain: This endpoint provides the wallet addresses needed by /copy-trader-trades. You MUST call this endpoint FIRST to get trader addresses — do NOT guess or hardcode addresses.
🚫 Self-copy guard: Never use your own user_wallet from /user-details as a copy-trader address.
# Get all traders (OpenClaw + Leaderboard)
curl -L -X GET "${MAXXIT_API_URL}/api/lazy-trading/programmatic/copy-traders" \
-H "X-API-KEY: ${MAXXIT_API_KEY}"# Get only OpenClaw Traders (prioritized)
curl -L -X GET "${MAXXIT_API_URL}/api/lazy-trading/programmatic/copy-traders?source=openclaw" \
-H "X-API-KEY: ${MAXXIT_API_KEY}"# Get only Leaderboard traders with filters
curl -L -X GET "${MAXXIT_API_URL}/api/lazy-trading/programmatic/copy-traders?source=leaderboard&minImpactFactor=50&minTrades=100" \
-H "X-API-KEY: ${MAXXIT_API_KEY}"
Query Parameters:
Parameter
Type
Default
Description
source
string
all
openclaw (OpenClaw agents only), leaderboard (top traders only), all (both)
Next step: After reviewing the trades, use /open-position to open a similar position. You'll need your own agentAddress and userAddress from /user-details.
Signal Format Examples
The lazy trading system processes natural language trading signals. Here are examples:
Opening Positions
"Long ETH with 5x leverage, entry at 3200"
"Short BTC 10x, TP 60000, SL 68000"
"Buy 100 USDC worth of ETH perpetual"
With Risk Management
"Long SOL 3x leverage, entry 150, take profit 180, stop loss 140"
"Short AVAX 5x, risk 2% of portfolio"
Closing Positions
"Close ETH long position"
"Take profit on BTC short"
Complete Workflow Examples
These are the mandatory step-by-step workflows for common trading operations. Follow these exactly.
Workflow 1: Opening a New Position (Full Flow)
Step 1: GET /user-details
→ Extract: user_wallet (→ userAddress), ostium_agent_address (→ agentAddress)
→ Cache these for the session
Step 2: GET /symbols
→ Verify the user's requested token is available on Ostium
→ Extract exact symbol string and maxLeverage
→ Convert pair format to market token for /open-position:
"ETH/USD" -> "ETH"
Step 3: POST /api/lazy-trading/research (or GET /market-data or GET /price for current context)
→ Get trade context: market structure, momentum, support/resistance, catalysts, and current price
→ Present this data to the user:
"BTC is trading around $95,000 with bullish structure and clear support/resistance levels.
Do you want to proceed?"
Step 4: ASK the user for trade parameters
→ "Please confirm: collateral (USDC), leverage, long or short?"
→ "Would you like to set TP and SL? If so, what percentages?"
→ Wait for explicit user confirmation before proceeding
Step 5: POST /open-position
→ Use agentAddress and userAddress from Step 1
→ Use market, side, collateral, leverage from Step 4
→ IMPORTANT: Pass market as base token only (e.g. ETH), not pair format (ETH/USD)
→ SAVE the response: actualTradeIndex and entryPrice
Step 6 (if user wants TP/SL): POST /set-take-profit and/or POST /set-stop-loss
→ Use tradeIndex = actualTradeIndex from Step 5
→ Use entryPrice from Step 5
→ For pairIndex, use the symbol id from Step 2 or call /positions
→ Use takeProfitPercent/stopLossPercent from Step 4
Step 7: ASK — "Would you like to list this trade as alpha on the marketplace?"
→ If user says NO → Done.
→ If user says YES → Continue to Step 8.
→ Also ask: "What price in USDC would you like to charge?" (e.g. 5 USDC)
Step 8: POST /alpha/generate-proof
→ Body: { "tradeId": "{tradeId from Step 5}", "autoProcess": false }
→ tradeId comes from the /open-position response
→ autoProcess: false queues the proof for the worker (~3-5 min)
→ SAVE: proofId from the response
Step 9: Wait for proof verification
→ If autoProcess was true and response has status: "VERIFIED" → go to Step 10
→ If autoProcess was false or status is still PENDING/PROVING:
Poll GET /alpha/proof-status?proofId={proofId} every 10 seconds
→ Wait until status === "VERIFIED"
→ If status === "FAILED" → inform the user and stop
Step 10: POST /alpha/flag
→ Body: {
"proofId": "{proofId from Step 8}",
"priceUsdc": {price from Step 7},
"token": "{market from Step 5, e.g. ETH}",
"side": "{side from Step 5, e.g. long}",
"leverage": {leverage from Step 5}
}
→ Show user the response: listingId, tradeId, proofMetrics
→ "Your trade is now listed as alpha! Listing ID: {listingId}"
Workflow 2: Closing an Existing Position
Step 1: GET /user-details
→ Extract: user_wallet, ostium_agent_address
Step 2: POST /positions (address = user_wallet from Step 1)
→ List all open positions
→ Present them to the user if multiple: "You have 3 open positions: BTC long, ETH short, SOL long. Which one do you want to close?"
→ Extract the tradeIndex for the position to close
Step 3: POST /close-position
→ Use agentAddress and userAddress from Step 1
→ Use market and actualTradeIndex from Step 2
→ Show the user the closePnl from the response
Workflow 3: Setting TP/SL on an Existing Position
Step 1: GET /user-details
→ Extract: user_wallet, ostium_agent_address
Step 2: POST /positions (address = user_wallet from Step 1)
→ Find the target position
→ Extract: tradeIndex, entryPrice, pairIndex, side
Step 3: ASK the user
→ "Position: BTC long at $95,000. Current TP: none, SL: $85,500."
→ "What TP% and SL% would you like to set?"
Step 4: POST /set-take-profit and/or POST /set-stop-loss
→ Use ALL values from Steps 1-3 — NEVER guess any of them
Workflow 4: Checking Portfolio & Market Overview
Step 1: GET /user-details
→ Extract: user_wallet
Step 2: POST /balance (address = user_wallet)
→ Show the user their USDC and ETH balances
Step 3: POST /positions (address = user_wallet)
→ Show all open positions with PnL details
Step 4 (optional): GET /market-data
→ Show market conditions for tokens they hold
Workflow 5: Copy-Trading Another OpenClaw Agent (Full Flow)
Step 1: GET /copy-traders?source=openclaw
→ Discover other OpenClaw Trader agents
→ Extract: creatorWallet from the trader you want to copy
→ Exclude your own wallet (`/user-details.user_wallet`) if it appears
→ IMPORTANT: This is a REQUIRED first step — you cannot call
/copy-trader-trades without an address from this endpoint
Step 2: GET /copy-trader-trades?address={creatorWallet}
→ Fetch recent trades for that trader from the Ostium subgraph
→ Review: side (LONG/SHORT), tokenSymbol, leverage, collateral, entry price
→ Decide: "Should I copy this trade?"
→ DEPENDENCY: The address param comes from Step 1 (creatorWallet or walletAddress)
Step 3: GET /user-details
→ Get YOUR OWN userAddress (user_wallet) and agentAddress (ostium_agent_address)
→ These are needed to execute your own trade
Step 4: POST /open-position
→ Mirror the trade from Step 2 using your own addresses from Step 3:
- market = tokenSymbol from the copied trade
- side = side from the copied trade (LONG/SHORT → long/short)
- collateral = decide based on your own risk tolerance
- leverage = match the copied trader's leverage or adjust
→ SAVE: actualTradeIndex and entryPrice from response
Step 5 (optional): POST /set-take-profit and/or POST /set-stop-loss
→ Use actualTradeIndex and entryPrice from Step 4
→ Match the copied trader's TP/SL ratios or set your own
Maxxit provides specialized scripts for different market conditions. These scripts require dynamic parameters passed via command line.
Execution Policy
Dynamic Arguments: Scripts MUST be invoked with --symbol and --venue.
Agent Responsibility: If the user asks to start a strategy but does not provide the symbol (e.g., "BTC/USD") or the venue (e.g., "AVANTIS"), the agent MUST ask the user for the missing information before executing the script.
Example Command: python taker-strategy.py --symbol BTC/USD --venue AVANTIS
Logic Summary: Monitors the "Taker Buy Ratio" on Binance. When aggressive buyers (takers) dominate sellers beyond a threshold (0.60), it signals a high-conviction momentum move.
Best For: Capturing rapid price changes in high-volume environments (Active Scalping).
Logic Summary: Combines RSI (extreme oversold/overbought) with Bollinger Band touches. It identifies "exhaustion" points where the price is likely to bounce back to the average.
Best For: Range-bound or sideways markets where there is no clear trend.
Logic Summary: Enters a trade only when price breaks out of a standard deviation channel (Bollinger Bands) and volatility (ATR) is increasing. This filters out "fake" breakouts.
Best For: Catching the start of a strong trend after a period of consolidation.
Logic Summary: Uses Volume Weighted Average Price (VWAP) combined with a 20 EMA. A "Long" is triggered when price is above both the VWAP and the EMA, signaling that both volume and time-weighted momentum are positive.
Best For: Intraday momentum trading and confirming trend strength with volume.
Aster DEX (BNB Chain) Endpoints
Aster DEX is a perpetual futures exchange on BNB Chain. Use Aster endpoints when the user wants to trade on BNB Chain. The Aster API uses API Key + Secret authentication (stored server-side) — you do NOT need agentAddress. You only need userAddress from /user-details.
Venue Selection
Venue
Chain
Symbol Format
Auth Required
When to Use
Ostium
Arbitrum (mainnet by default, testnet on explicit request)
BTC, ETH
agentAddress + userAddress
Default for most trades
Aster
BNB Chain (testnet only)
BTCUSDT, ETHUSDT
userAddress only
When user specifies BNB Chain or Aster
Avantis
Base (mainnet only)
Base token for orders (e.g. BTC), pair format in symbols/positions (e.g. BTC/USD)
agentAddress + userAddress
When user specifies Base chain or Avantis
Network behavior rule: Do not ask users to choose mainnet/testnet for these venues by default. Ostium uses mainnet unless the user explicitly asks for testnet / Arbitrum Sepolia. Aster is fixed to testnet, and Avantis is fixed to Base mainnet.
How to check if Aster is configured: In the /user-details response, aster_configured: true means the user has set up Aster API keys. If false, direct them to set up Aster at maxxit.ai/openclaw.
Aster Symbols
Aster uses Binance-style symbol format: BTCUSDT, ETHUSDT, etc. The API auto-appends USDT if you pass just BTC.
curl -L -X GET "${MAXXIT_API_URL}/api/lazy-trading/programmatic/aster/symbols" \
-H "X-API-KEY: ${MAXXIT_API_KEY}"
{"userAddress":"0x...",// REQUIRED — from /user-details → user_wallet"symbol":"BTC",// REQUIRED — token or full symbol (BTC or BTCUSDT)"limit":100,// Optional — default depends on exchange (max 1000)"orderId":12345,// Optional — fetch from this orderId onward"startTime":1709251200000,// Optional — ms timestamp"endTime":1709856000000// Optional — ms timestamp}
POST /api/lazy-trading/programmatic/aster/history now proxies to Aster /fapi/v3/allOrders.
Use this endpoint when users ask for "all old trades/orders", "order history", or "past orders" on Aster.
Aster Open Position
📋 LLM Pre-Call Checklist — Ask the user these questions before calling this endpoint:
Symbol: "Which token do you want to trade?" (e.g. BTC, ETH, SOL)
Side: "Long or short?"
Quantity: "How much [TOKEN] do you want to trade?" — get the answer in base asset units (e.g. 0.01 BTC, 0.5 ETH).
Leverage: "What leverage? (e.g. 10x)"
Order type: "Market order or limit order?" (default: MARKET). If LIMIT, also ask for the limit price.
Aster requires quantity (base asset) for open-position. Do not use collateral.NEVER call this endpoint without a confirmed quantity in base asset units.
{"userAddress":"0x...",// REQUIRED — from /user-details → user_wallet. NEVER guess."symbol":"BTC",// REQUIRED — Token name or full symbol (BTCUSDT). ASK the user."side":"long",// REQUIRED — "long" or "short". ASK the user."quantity":0.01,// REQUIRED — Position size in BASE asset (e.g. 0.01 BTC). ASK the user."leverage":10,// Optional — Leverage multiplier. ASK the user."type":"MARKET",// Optional — "MARKET" (default) or "LIMIT". ASK the user."price":95000// Required only for LIMIT orders. ASK the user if type is LIMIT.}
⚠️ IMPORTANT:quantity must always be specified in the base asset (e.g. 0.01 for 0.01 BTC).
If the user provides a USDT/collateral amount, ask them to provide the exact token quantity instead.
Do not convert collateral to quantity in this workflow.
Response (IMPORTANT — save these values):
{"success":true,"orderId":12345678,"symbol":"BTCUSDT","side":"BUY","status":"FILLED","avgPrice":"95000.50","executedQty":"0.010","message":"Position opened: long BTCUSDT"}
User specifies in base asset units (e.g. 0.01 BTC)
User input (required). If user provides USDT/collateral amount, ask for quantity instead. Do not calculate in the workflow.
leverage
User specifies
User input
entryPrice
/aster/positions → entryPrice
From position data
stopPrice
User specifies or calculated from percent
User input or calculated
Aster Workflow: Open Position on BNB Chain
Step 1: GET /user-details
→ Extract: user_wallet
→ Check: aster_configured == true (if false, tell user to set up Aster at maxxit.ai/openclaw)
Step 2: GET /aster/symbols
→ Verify the token is available on Aster
Step 3: GET /aster/price?token=BTC
→ Get current price, present to user
Step 4: ASK the user for ALL trade parameters
→ "Which token?" (e.g. BTC, ETH, SOL)
→ "Long or short?"
→ "How much [TOKEN] do you want to buy/sell?" — collect answer in BASE asset units (e.g. 0.01 BTC)
• If user gives a USDT/collateral amount, ask them to provide token quantity instead.
→ "Leverage? (e.g. 10x)"
→ "Market or limit order?" — if LIMIT, also ask for the limit price
Step 5: POST /aster/open-position
→ Use userAddress from Step 1
→ Use symbol, side, quantity (base asset), leverage from Step 4
→ SAVE orderId and avgPrice from response
Step 6 (if user wants TP/SL): POST /aster/set-take-profit and/or POST /aster/set-stop-loss
→ Use entryPrice = avgPrice from Step 5
→ Use side from Step 4
→ Ask user for takeProfitPercent / stopLossPercent (or exact stopPrice)
Aster Workflow: Close Position
Step 1: GET /user-details → Extract user_wallet
Step 2: POST /aster/positions (userAddress = user_wallet)
→ Show positions to user, let them pick which to close
Step 3: POST /aster/close-position
→ Pass userAddress and symbol
→ Omit quantity to close full position
Avantis DEX (Base Chain) Endpoints
Avantis DEX is a perpetual futures exchange on Base chain. Use Avantis endpoints when the user wants to trade on Base. Avantis uses delegation-based auth (same pattern as Ostium) — you need both agentAddress and userAddress from /user-details.
How to check if Avantis is configured: Use /user-details and check deployment.enabled_venues. If it includes "AVANTIS", Avantis is enabled for the current deployment. If not, direct the user to enable Avantis at maxxit.ai/openclaw.
Avantis Symbols
Avantis symbols are returned in pair format (e.g. BTC/USD, ETH/USD). The API endpoint maps the service's /markets route.
curl -L -X GET "${MAXXIT_API_URL}/api/lazy-trading/programmatic/avantis/symbols" \
-H "X-API-KEY: ${MAXXIT_API_KEY}"
{"agentAddress":"0x...",// REQUIRED — from /user-details → ostium_agent_address (shared wallet). NEVER guess."userAddress":"0x...",// REQUIRED — from /user-details → user_wallet. NEVER guess."market":"BTC",// REQUIRED — Base token only (e.g. "ETH", not "ETH/USD")"side":"long",// REQUIRED — "long" or "short". ASK the user."collateral":100,// REQUIRED — Collateral in USDC. ASK the user."leverage":10,// Optional (default: 10). ASK the user."takeProfitPercent":0.30,// Optional — set TP at open. ASK the user."stopLossPercent":0.10// Optional — set SL at open. ASK the user.}
Response:
{"success":true,"txHash":"0x...","actualTradeIndex":5,"entryPrice":95000.0,"slSet":true,"tpSet":true,"message":"Trade submitted on Avantis","result":{"market":"BTC","side":"long","collateral":100,"leverage":10,"slConfigured":true,"tpConfigured":true,"tpPrice":123500.0,"slPrice":85500.0}}
{"agentAddress":"0x...",// REQUIRED — from /user-details. NEVER guess."userAddress":"0x...",// REQUIRED — from /user-details. NEVER guess."market":"BTC",// REQUIRED — Token symbol"tradeId":"0:2",// Optional — preferred composite ID from /avantis/positions (pairIndex:tradeIndex)"actualTradeIndex":2// Recommended — from /avantis/positions → tradeIndex}
Avantis Update SL/TP
TP/SL can be set at open time (via takeProfitPercent/stopLossPercent in open-position), or updated after opening using this endpoint.
{"agentAddress":"0x...",// REQUIRED — from /user-details. NEVER guess."userAddress":"0x...",// REQUIRED — from /user-details. NEVER guess."market":"BTC",// REQUIRED — Token symbol"tradeIndex":0,// Optional — specific trade index from /avantis/positions"takeProfitPrice":100000,// Absolute TP price (use this OR takeProfitPercent)"stopLossPrice":80000,// Absolute SL price (use this OR stopLossPercent)"takeProfitPercent":0.30,// TP as % from entry (0.30 = 30%). Use this OR takeProfitPrice."stopLossPercent":0.10// SL as % from entry (0.10 = 10%). Use this OR stopLossPrice.}
{"venue":"AVANTIS",// REQUIRED for Avantis via /history"userAddress":"0x...",// REQUIRED — the trader's wallet address"agentAddress":"0x...",// Alternative to userAddress"count":50// Optional — max results (default: 50)}
Step 1: GET /user-details
→ Extract: user_wallet, ostium_agent_address (shared agent wallet)
→ Check: deployment.enabled_venues includes AVANTIS (if not, tell user to enable Avantis at maxxit.ai/openclaw)
Step 2: GET /avantis/symbols
→ Verify the token is available on Avantis
Step 3: ASK the user for ALL trade parameters
→ "Which token?" (e.g. BTC, ETH)
→ "Long or short?"
→ "How much USDC collateral?"
→ "Leverage? (e.g. 10x)"
→ "Would you like to set TP/SL? If so, what percentages?"
Step 4: POST /avantis/open-position
→ Use agentAddress and userAddress from Step 1
→ Use market, side, collateral, leverage from Step 3
→ SAVE: tradeIndex and entryPrice from response
Avantis Workflow: Close Position
Step 1: GET /user-details → Extract user_wallet, ostium_agent_address (shared agent wallet)
Step 2: POST /avantis/positions (userAddress + agentAddress)
→ Show positions to user, let them pick which to close
→ Extract tradeIndex
Step 3: POST /avantis/close-position
→ Pass agentAddress, userAddress, market
→ Optionally pass actualTradeIndex
Zerodha (Indian Stocks) Endpoints
Zerodha is an Indian stock broker supporting equities on NSE and BSE. Use Zerodha endpoints when the user wants to trade Indian stocks, mentions NSE/BSE, or says "Indian stocks", "equities", or "Zerodha".
When to Use Zerodha
User wants to trade Indian stocks on NSE or BSE
User mentions "Indian stocks", "equities", "NSE", "BSE", "Zerodha", or "Kite"
User asks about their Zerodha portfolio, holdings, or positions
User wants to place, modify, or cancel orders on Indian exchanges
User wants to fetch instrument lists or market data for Indian equities
Currency Convention
In Indian-market conversations, present prices, holdings, portfolio values, targets, and stop losses in ₹
Prefer ₹ in normal user-facing replies
Do not default to USD when the context is Zerodha, NSE, BSE, or Indian equities
TP/SL and GTT Guidance
When the user wants to set a take profit or stop loss for a Zerodha position, confirm whether they want to place a GTT order.
Explain GTT briefly when needed: GTT (Good Till Triggered) is a Zerodha trigger order that stays active until the trigger condition is hit or the user cancels it, and it is commonly used to automate target and stop-loss execution for cash-market positions.
If the user explicitly asks for GTT, use the Zerodha GTT flow directly.
If the user asks for TP/SL but does not mention GTT, ask first instead of assuming they want a regular order or a GTT trigger.
If authenticated: false or expired: true, tell the user:
"Your Zerodha session has expired. Please re-authenticate on the OpenClaw page at maxxit.ai/openclaw."
Important: All Zerodha requests must include:
X-API-KEY header (normal Maxxit auth)
X-KITE-API-KEY header
X-KITE-ACCESS-TOKEN header
For wallet identity, call /user-details first and use user_wallet. Zerodha does not require ostium_agent_address or lazy_trading_ready: true.
Zerodha Endpoints
All base path: ${MAXXIT_API_URL}/api/lazy-trading/programmatic/zerodha/
GET /zerodha/login
Generate a Zerodha login URL for the user.
Important:
Do not tell the user to open a bare Kite URL like https://kite.zerodha.com/connect/login?api_key=...&v=3.
The login flow must preserve the Maxxit user context so the callback can store the Zerodha session against the correct wallet.
When the user asks for the login link, give them the login_url exactly as returned by the Maxxit API. That URL may use the Kite domain, but it must include the Maxxit user handoff in redirect_params.
Preferred flow:
Call GET /user-details to resolve user_wallet.
Call GET /zerodha/login with X-API-KEY.
Send the returned login_url to the user.
Do not manually construct a ?userWallet=<wallet>&redirect=1 URL in the normal skill flow.
If you present the returned login_url, only use it if it includes the user handoff, e.g. redirect_params=userWallet%3D....
curl -L -X GET "${MAXXIT_API_URL}/api/lazy-trading/programmatic/zerodha/login" \
-H "X-API-KEY: ${MAXXIT_API_KEY}"
Response:
{"success":true,"login_url":"https://kite.zerodha.com/connect/login?api_key=5wh5hi6ky7y8s6g4&v=3&redirect_params=userWallet%3D0x796a837c78326ba693847deebd7811d6b6854c56","message":"Open the login_url in your browser to authenticate with Zerodha."}
GET /zerodha/session
Check Zerodha session status. Returns authenticated: true if session is valid.
How to explain Zerodha order fields to a normal trader:
Field
Value
Meaning for a trader
variety
regular
Standard exchange order placed during market hours. Use this by default unless the user asks for something more specific.
variety
amo
After Market Order. Place the order outside market hours so it gets queued for the next session.
variety
co
Cover Order. An intraday order paired with a compulsory stop loss. Use only if the user specifically wants a leveraged intraday order with a built-in risk stop.
variety
iceberg
Iceberg Order. Splits one large order into smaller legs so the full size is not sent at once. Useful for quantities above freeze limits or to reduce visible impact.
variety
auction
Auction Order. Used for exchange auction sessions, not for normal cash-market trading. Only use when the user explicitly wants auction participation.
order_type
MARKET
Buy or sell immediately at the best available price. Prioritizes execution over exact price.
order_type
LIMIT
Execute only at the specified price or better. Prioritizes price control over certainty of fill.
order_type
SL
Stop-loss limit order. Once the trigger is hit, Zerodha places a limit order. Gives price control, but the order may remain unfilled in fast moves.
order_type
SL-M
Stop-loss market order. Once the trigger is hit, Zerodha places a market order. Better for ensuring exit, but fill price can slip.
product
CNC
Cash and Carry. Standard delivery equity order for shares you want to hold, typically in NSE/BSE cash markets.
product
NRML
Normal order type commonly used for futures and options carry-forward positions, and other non-intraday derivatives exposure.
product
MIS
Margin Intraday Squareoff. Intraday-only product, typically for futures/options or same-day cash trading where the position is not meant to be carried overnight.
product
MTF
Margin Trading Facility. Broker-funded carry-forward equity position. Use only if the user explicitly asks for MTF and their account supports it.
validity
DAY
Order stays active for the current trading session unless filled or cancelled.
validity
IOC
Immediate or Cancel. Whatever can fill immediately will execute; the rest is cancelled right away.
validity
TTL
Time to Live. The order stays active only for the specified number of minutes, using validity_ttl.
market_protection
0
No market protection. Default behavior.
market_protection
0 - 100
Custom market-protection percentage for market and stop-market style execution. Helps cap how far execution can chase price. Example: 2 means 2%.
market_protection
-1
Automatic market protection chosen by the system based on Zerodha rules.
autoslice
true
Automatically split a large order into multiple slices when quantity is above freeze limits.
autoslice
false
Do not auto-slice. Default behavior.
Extra order parameters supported by this route:
validity_ttl: number of minutes when validity is TTL
iceberg_legs: number of child legs for an iceberg order
iceberg_quantity: quantity per iceberg leg
auction_number: required when placing auction orders
market_protection: market protection setting for eligible orders
autoslice: whether Zerodha should automatically slice oversized orders
Agent guidance:
Default to variety: regular, validity: DAY, and either MARKET or LIMIT based on what the user asks.
Do not choose CO, iceberg, auction, MTF, TTL, market_protection, or autoslice unless the user explicitly wants that behavior or it is clearly necessary for the order they described.
If the user asks in plain English, translate the intent into these fields and explain the tradeoff briefly before placing the order.
