· okx integration

AlgoVault has 89.4%+ PFE Win Rate across 56,375+ signal calls, each Merkle-anchored on Base L2 (verifiable at ).

AlgoVault × OKX — Build Verifiable AI Trading Agents

AlgoVault MCP gives your agent a composite verdict in one call — direction, confidence, regime, and cross-venue funding/sentiment context — backed by a publicly auditable record anchored to Base L2. Pair it with OKX’s Agent Trade Kit and your agent has both the analytics brain and the execution venue.

Provenance: Official OKX @okx_ai/[email protected] (npm) + github.com/okx/agent-trade-kit (default branch master, description “OKX trading MCP server — connect AI agents to spot, swap, futures, options & grid bots via the Model Context Protocol”). Demo execution runs against OKX’s simulated-trading environment via the --demo CLI flag (verbatim from docs/cli-reference.md: --demo | Use simulated trading (demo) mode). Verified 2026-04-25.

What you’ll build (90s read)

A runnable Node.js agent that:

  1. Calls AlgoVault MCP for get_market_regime across BTC, ETH, and SOL on the 4h timeframe.
  2. Identifies any asset whose regime is RANGING (a precondition for grid-bot strategies).
  3. Hands the candidate set to the OKX Agent Trade Kit, which would create a small grid-bot order in OKX demo mode when the agent’s pre-configured policy fires.

The whole loop runs against OKX’s simulated-trading environment (the --demo flag’s underlying behavior; equivalent REST surface via the x-simulated-trading: 1 header). Zero real-money risk in any code path.

Prerequisites (4 items)

  1. Node.js ≥ 22 (node --version to check).
  2. AlgoVault skills plugin installed:
    claude plugin install AlgoVaultLabs/algovault-skills
  3. OKX demo trading account — sign in at https://www.okx.com → toggle to Demo Trading (top-right account menu). Generate API key + secret + passphrase from the demo console (separate from mainnet keys).
  4. OKX Agent Trade Kit installed (recommended for the recipes; the demo runs without it via the public REST surface):
    npx -y @okx_ai/okx-trade-cli@latest setup --client claude-code --profile demo
    Then:
    npx -y @okx_ai/okx-trade-cli@latest --demo --modules market,spot,account

Demo: Multi-asset regime scan + grid-bot trigger (≤80 lines)

// examples/okx/demo.mjs (excerpt — see file for full source)
import { getAlgoVaultVerdict } from '../_shared/algovault-helper.mjs';

const MAINNET_BLOCKED = true;
const OKX_BASE = 'https://www.okx.com/api/v5';
const SIMULATED = { 'x-simulated-trading': '1' };  // OKX demo-trading header
const ASSETS = ['BTC', 'ETH', 'SOL'];

if (process.env.OKX_DEMO !== 'true') {
  throw new Error('OKX_DEMO=true required — demo refuses to run against mainnet.');
}

console.log('=== DEMO MODE ===');
console.log(`[OKX_DEMO=true | --demo equivalent header x-simulated-trading: 1]`);

// 1. AlgoVault regime read across BTC/ETH/SOL on 4h
const verdicts = await Promise.all(
  ASSETS.map((coin) => getAlgoVaultVerdict({ coin, timeframe: '4h' }))
);

// 2. Identify RANGING regimes (grid-bot precondition)
const rangingAssets = verdicts.filter((v) => v.regime === 'RANGING');

// 3. Probe OKX simulated-trading instrument (proves --demo path works)
//    Real grid-bot creation uses POST /api/v5/tradingBot/grid/order-algo

console.log('=== NO REAL ORDERS PLACED ===');

Run it:

OKX_DEMO=true node examples/okx/demo.mjs

Expected output (last 4 lines):

[verdicts | BTC=... ETH=... SOL=... ranging_count=...]
[okx demo | spot/instruments HTTP 200 | host=www.okx.com (x-simulated-trading: 1)]
[grid-bot | SKIPPED — see README for authenticated POST /tradingBot/grid/order-algo]
=== NO REAL ORDERS PLACED ===

Walkthrough (line-by-line — neutral narration)

The script does three things in order:

  1. Fetch AlgoVault regime reads in parallel. The shared helper opens an MCP session against https://api.algovault.com/mcp, calls get_trade_signal for BTC / ETH / SOL on the 4h timeframe (each call returns regime alongside the signal/confidence), and returns the parsed payloads. Free tier covers 20 calls/day per IP — three regime reads per scan is ample for development.

  2. Apply the agent’s policy. The agent’s grid-bot policy fires only when at least one asset in the scan returns a RANGING regime (grid bots only profit in range-bound markets — running them in TRENDING_* or VOLATILE regimes destroys capital). The policy lives in your code — AlgoVault returns the regime classifications; the agent decides.

  3. Probe the OKX simulated-trading surface. The script issues a GET /api/v5/public/instruments?instType=SPOT&instId=BTC-USDT against www.okx.com with the x-simulated-trading: 1 header (the REST equivalent of the OKX MCP CLI’s --demo flag). On success OKX returns code: "0" and a full instrument-detail payload. The === NO REAL ORDERS PLACED === banner closes every run regardless of the policy outcome. The script aborts immediately if OKX_DEMO is not true — a hard guard against accidentally running against mainnet.

3 Recipes

Recipe 1 — Multi-asset regime scan + grid-bot trigger

This is the recipe examples/okx/demo.mjs implements. The agent calls AlgoVault get_market_regime (or get_trade_signal which carries regime as a field) across BTC, ETH, and SOL on the 4h timeframe. For any asset whose regime is RANGING, the agent’s pre-configured policy hands the candidate to the OKX Agent Trade Kit’s grid-bot creator (POST /api/v5/tradingBot/grid/order-algo) in OKX demo mode. The grid bounds, leverage, and grid count come from the agent’s own configuration — AlgoVault returns the regime; the agent tunes the bot.

Recipe 2 — Funding-arb pair on options

The agent calls AlgoVault scan_funding_arb to retrieve the top funding-spread opportunities. When the top spread exceeds 12 bps and OKX is one of the two indicated venues, the agent’s pre-configured policy can route the long leg through OKX options (POST /api/v5/trade/order with instType=OPTION) in OKX demo mode and the short leg through the counterparty venue. Multi-venue execution is out of scope for this single-exchange tutorial — readers are advised to extend the demo with their own cross-venue routing layer once the policy thresholds are dialled in.

Recipe 3 — Risk-gated entry with confidence floor

The agent calls AlgoVault get_trade_signal for the asset of interest. The agent’s pre-configured policy fires only when the returned confidence exceeds 75% — a deliberately tight floor that filters out mid-conviction signals. On execution, a small spot order goes to OKX demo via POST /api/v5/trade/order with instType=SPOT. The threshold of 75% is the agent’s choice, not AlgoVault’s recommendation — you tune it to your strategy’s risk appetite.

⚠️ Production setup (real-money)

The demo above runs in OKX demo trading only. Real-money setup requires:

  • KYC completion on OKX.
  • Production API keys with only the permissions your agent needs (no withdrawals, IP-allowlisted).
  • Risk controls: per-order size cap, daily loss limit, kill switch.
  • Position monitoring: a separate agent or watchdog that tracks open positions independently.

See examples/okx/README.md for the full real-money checklist + the OKX Agent Trade Kit production install via npx -y @okx_ai/okx-trade-cli@latest setup --client claude-code --profile live. AlgoVault provides analytics; your agent and your risk policy decide what (if anything) to execute.

Why AlgoVault? (closing — MOAT recap)

  • Composite verdict, not raw indicators. One JSON response replaces 26-indicator vote-counting.
  • Cross-venue intelligence. Funding spreads, regime, and sentiment fused across 5 exchanges — not derivable from any single-venue API.
  • Publicly verified. Every signal anchored to Base L2 via Merkle proof. Verify before you subscribe.

89.4% PFE Win Rate · 56,375+ calls · 16+ on-chain batches → view live track record

Install

claude plugin install AlgoVaultLabs/algovault-skills

Once installed, every Skill in the pack is one-line invokable from Claude Code, Cowork, or any MCP-compatible client.

Tutorial © AlgoVault Labs · MIT licensed · Provenance verified 2026-04-25 · OKX @okx_ai/[email protected] (npm) + okx/agent-trade-kit (GitHub)