StandX Agent Toolkit

OpenClaw First. AI Agent Native. Trading Ecosystem Ready.

StandX Agent Toolkit is a CLI designed for the AI Trading era—OpenClaw First, yet universally adaptable to any AI Agent that can execute commands.

We believe the future of trading is conversational. Your agent should trade as naturally as it chats. No complex APIs, no boilerplate—just intent to execution.

┌─────────────────────────────────────────────────────────────────┐
│                                                                 │
│   You: "Check my BTC position"                                  │
│   ↓                                                             │
│   OpenClaw → StandX CLI → StandX API                            │
│   ↓                                                             │
│   You: "Long 0.1 BTC, stop loss at $62k"                        │
│   ↓                                                             │
│   ✅ Order executed in seconds                                  │
│                                                                 │
│   Your agent now trades as naturally as it converses.           │
│                                                                 │
└─────────────────────────────────────────────────────────────────┘

---

🎯 Why StandX Agent Toolkit?

The Problem

You have an AI Agent (OpenClaw, Claude, AutoGPT, etc.). You want it to trade. But:

• ❌ Traditional trading tools are built for humans clicking buttons
• ❌ APIs require complex integration and parsing
• ❌ No bridge between natural language and execution

The Solution

Agent-First Design—structured output, non-interactive, composable:

Feature	Traditional Tools	StandX Agent Toolkit
Built For	Human traders	AI Agents
OpenClaw Integration	Custom code	Works out of the box
Output	Pretty tables	Structured JSON
Errors	Text to parse	Machine-readable
Workflow	Interactive prompts	100% scriptable
Other Agents	Not supported	CLI = Universal

---

🚀 Quick Start

1. Install

Option 1: One-line Installer (Recommended)

# macOS (Apple Silicon) / Linux (x86_64 & ARM64)
curl -sSL https://raw.githubusercontent.com/wjllance/standx-cli/main/install.sh | sh

Option 2: Homebrew (macOS)

brew tap wjllance/standx-cli
brew install standx-cli

Option 3: Build from Source

cargo install standx-cli

2. Configure

StandX CLI requires authentication for most operations. You need:

1. JWT Token (required) - For reading account data
2. Ed25519 Private Key (optional, but recommended) - For trading operations

Get Credentials

Visit https://standx.com/user/session to generate:

• JWT Token (valid for 7 days)
• Ed25519 Private Key (Base58 encoded)

Login Methods

Interactive (Recommended for first-time setup):

standx auth login --interactive

Command line (for scripts/agents):

standx auth login \
  --token "$STANDX_JWT" \
  --private-key "$STANDX_PRIVATE_KEY"

From files:

standx auth login \
  --token-file ~/.standx_token \
  --key-file ~/.standx_key

Environment variables (auto-detected):

export STANDX_JWT="your_jwt_token"
export STANDX_PRIVATE_KEY="your_private_key"

Check Authentication Status

standx auth status

Example output:

✅ Authenticated
   Token expires at: 2024-02-02T09:56:07Z
   Remaining: 167 hours

Logout

standx auth logout

Permission Requirements

Operation	JWT Token	Private Key
Market data (ticker, depth)	❌ No	❌ No
Account info (balances, positions)	✅ Yes	❌ No
View orders & trades	✅ Yes	❌ No
Create/cancel orders	✅ Yes	✅ Yes
Change leverage	✅ Yes	✅ Yes
Margin operations	✅ Yes	✅ Yes

Note: Trading operations require the Ed25519 private key for request signing. If you only provide the JWT token, you'll see: ⚠️ No private key provided - trading operations will be unavailable

For detailed authentication documentation, see docs/02-authentication.md.

3. Use With Your Agent

OpenClaw (Native)

You: What's the BTC price?
OpenClaw: [executes: standx market ticker BTC-USD --output json]
          BTC is trading at $65,000 (+2.3% today)

You: Buy 0.1 BTC at market price
OpenClaw: [executes: standx order create BTC-USD buy market --qty 0.1]
          ✅ Market order executed
          Bought 0.1 BTC at $65,001

Claude / Cursor / Any CLI-capable Agent

# Same commands work everywhere
import subprocess

result = subprocess.run(
    ["standx", "market", "ticker", "BTC-USD", "--output", "json"],
    capture_output=True
)
data = json.loads(result.stdout)

---

🛠️ Integration Patterns

Pattern 1: OpenClaw Native (Recommended)

OpenClaw calls StandX CLI directly via exec:

# In OpenClaw
result = await exec("standx market ticker BTC-USD --output json")
price_data = json.loads(result.stdout)
# Agent parses and responds naturally

Best for: OpenClaw users who want seamless conversation-to-trading

Pattern 2: Universal CLI

Any AI Agent that can execute shell commands:

# LangChain
from langchain.tools import ShellTool

tool = ShellTool()
result = tool.run("standx account balances --output json")

# AutoGPT
# Add to skills
os.system("standx order create BTC-USD buy market --qty 0.1")

Best for: Multi-platform agents, custom workflows

Pattern 3: Future MCP (Optional)

When you need richer tool definitions:

# Coming soon
standx mcp serve

Best for: Complex multi-step workflows across multiple services

---

📋 Command Reference

Market Data

# Price
standx market ticker BTC-USD --output json

# Order book
standx market depth BTC-USD --limit 10 --output json

# Recent trades
standx market trades BTC-USD --limit 20 --output json

# Funding rate
standx market funding BTC-USD --days 7 --output json

Account

# Balance
standx account balances --output json

# Positions
standx account positions --symbol BTC-USD --output json

# Open orders
standx account orders --symbol BTC-USD --output json

Trading

# Market order
standx order create BTC-USD buy market --qty 0.1

# Limit order
standx order create BTC-USD buy limit --qty 0.1 --price 64000

# With stop loss and take profit
standx order create BTC-USD buy limit --qty 0.1 --price 64000 \
  --sl-price 62000 --tp-price 68000

# Cancel
standx order cancel BTC-USD --order-id ord_xxx
standx order cancel-all BTC-USD

Dashboard

# Launch real-time trading dashboard
standx dashboard

# Watch specific symbols
standx dashboard --symbols BTC-USD,ETH-USD,SOL-USD

# Auto-refresh mode (updates every 5 seconds)
standx dashboard --watch

Leverage & Margin

# Get leverage
standx leverage get BTC-USD

# Set leverage
standx leverage set BTC-USD 10

# Get margin mode
standx margin mode BTC-USD

# Set margin mode
standx margin mode BTC-USD --set isolated

:### Trade History

# Get recent trades
standx trade history BTC-USD --from 1d

# With time range
standx trade history BTC-USD --from 2024-01-01 --to 2024-01-07

Portfolio

# Get portfolio summary
standx portfolio

# Verbose mode with more details
standx portfolio --verbose

# Auto-refresh mode
standx portfolio --watch

Streaming

# Real-time price stream
standx stream price BTC-USD

# Order book depth
standx stream depth BTC-USD --levels 5

# Public trades
standx stream trade BTC-USD

# Authenticated streams (requires login)
standx stream order      # Order updates
standx stream position   # Position updates
standx stream balance    # Balance updates
standx stream fills      # Fill updates

---

💡 Use Cases

1. Natural Language Trading (OpenClaw)

You: "I want to long ETH with 0.5 size, entry at 3500"
OpenClaw: "I'll place a limit buy order for 0.5 ETH at $3,500. 
           Current price is $3,480. Confirm?"
You: "Yes"
OpenClaw: "✅ Order placed. Order ID: ord_eth_xxx"

2. Automated Strategy (Any Agent)

# Grid trading bot
async def grid_trade():
    ticker = await exec("standx market ticker BTC-USD --output json")
    price = json.loads(ticker.stdout)["mark_price"]
    
    if price < lower_bound:
        await exec(f"standx order create BTC-USD buy limit --qty 0.01 --price {buy_price}")

3. Multi-Agent Coordination

# Risk monitoring agent
while True:
    positions = await exec("standx account positions --output json")
    # Alert if exposure too high
    
# Execution agent
await exec("standx order create ...")

---

🗺️ Roadmap

Phase 1: OpenClaw Excellence ✅ (Completed)

Goal: Best-in-class OpenClaw integration

• Structured JSON output
• Non-interactive mode
• Dashboard for real-time monitoring
• WebSocket streaming
• Complete trading commands (order, leverage, margin)
• --openclaw optimized defaults
• Session persistence
• Batch execution

Phase 2: Universal Agent Toolkit (Current)

Goal: Seamless experience across all AI Agents

• Comprehensive testing framework
• Portfolio PnL analysis
• Python SDK - pip install standx-agent
• Strategy templates (Grid, DCA, TWAP)
• Webhook callbacks
• MCP support (optional enhancement)

Phase 3: AI Trading Ecosystem (Future)

Goal: Define the standard for AI-native trading

• Multi-exchange abstraction
• Natural language strategy builder
• Agent marketplace
• Cross-agent coordination protocol

---

🤝 Comparison

Tool	OpenClaw	Other Agents	Learning Curve
StandX Agent Toolkit	🟢 Native	🟢 CLI = Universal	🟢 Low
Hummingbot	🔴 Complex	🔴 Complex	🔴 High
CCXT	🟡 Wrapper needed	🟡 Wrapper needed	🟡 Medium
Hyperliquid SDK	🟡 Integration needed	🟡 Integration needed	🟡 Medium

---

🛡️ Safety Features

• Structured errors - Agents can handle errors programmatically
• Dry-run mode - Test without execution
• Confirmation controls - --confirm / --no-confirm
• Rate limiting - Built-in protection

---

📝 Philosophy

OpenClaw First — We optimize for the best OpenClaw experience first.

Agent Native — Every design decision prioritizes machine consumption over human readability.

Ecosystem Ready — CLI is the universal interface. Works with any agent, today.

Future Proof — MCP, SDKs, and advanced features come later. The foundation is solid.

---

📜 License

MIT OR Apache-2.0

---

Built for the AI Trading era.

OpenClaw First. Agent Native. Ecosystem Ready.

---

💰 Buy Me Some Tokens

┌────────────────────────────────────────────────────────────┐
│                                                            │
│   🤖 Your AI agent made some gains?                        │
│                                                            │
│   💸 Buy it some oil (sponsor API tokens)                  │
│                                                            │
│   ┌────────────────────────────────────────────────┐      │
│   │  0xAb3D58779dFC50BC84caA796003ABE31b5296210   │      │
│   └────────────────────────────────────────────────┘      │
│                                                            │
│   ✨ Support ongoing development & maintenance ✨          │
│                                                            │
└────────────────────────────────────────────────────────────┘

EVM: 0xAb3D58779dFC50BC84caA796003ABE31b5296210

Every token counts. Even a gas fee is appreciated! ⛽🙏