mcpField
P
Protocol Banks

Protocol Banks

Remote
github.com/everest-anOther3Updated today
Repository

About Protocol Banks

Protocol Bank provides an AI trading product and programmatic trading API for agents using Hyperliquid. It is designed for wallet users and agent builders, with server-enforced spending policies, trading-only permissions, and operation audits.

Key features

  1. 01Automated perpetual-market trading
  2. 02Paper trading with real market data
  3. 03Per-order and daily risk limits
  4. 04Revocable trading-agent permissions
  5. 05Policy-enforced agent API
  6. 06Operation audit trail

Use cases

  1. 01Testing strategies with simulated funds
  2. 02Letting an agent trade within defined caps
  3. 03Pausing or revoking an agent's trading access
  4. 04Integrating agent trading through an API

MCP support

The MCP server lets MCP hosts such as Claude Desktop and Claude Code query and control the trading agent.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://protocolbanks.com/api/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

Protocol Bank — AI Automated Trading

Protocol Bank Logo

Your AI trades. You keep control.

Next.js TypeScript Go Prisma CI

English | 中文


Overview

Protocol Bank is an AI automated trading product. Users connect a wallet (MetaMask), fund a trading budget on Hyperliquid, approve a trading-only agent wallet, and an AI agent trades perpetual markets around the clock — with strict risk controls, full transparency, and instant revocation. A paper-trading mode runs on real market data with simulated money, so anyone can watch the agent work risk-free.

The same execution layer is exposed as a programmatic API for AI agents: one credential, one policy the server enforces (operations, per-call cap, daily cap, coin allowlist), and an audit trail per operation. See docs/AGENT_API.md for the reference and docs/AGENT_CASES.md for four worked integrations.

Core safety properties:

  • The AI can trade, never withdraw — the agent wallet holds trading-only permissions on Hyperliquid
  • Worst case is written on the screen — the trading wallet is the maximum possible loss
  • Guarded by construction — per-trade stop-loss, position caps, daily loss circuit breakers
  • Revocable anytime — pause, emergency stop, or revoke the agent on Hyperliquid

The Product

AreaDescription
AI Trading Cockpit (/trading)Post-login landing: balances, equity curve, plain-language AI activity feed, open positions, pause/stop/reset controls
Paper EngineReal Hyperliquid market data + simulated fills. Momentum (24h z-score) and funding-carry signals, 2.5% TP/SL, 15% position sizing, 5%/8% daily circuit breakers
Live ModeAgent wallet approval flow (EIP-712 approveAgent), AES-256-GCM key custody, real IOC order execution signed by the agent key, uncertain-order safety halt, manual approval mode (opt-in), per-user isolation schema
Wallet (/balances)Multi-chain balances and activity (merged from the legacy dashboard)
Business (collapsed)The legacy enterprise payment suite: batch payments, invoices, subscriptions, acquiring, vendors — available but no longer the hero

Architecture

┌──────────────────────────────────────────────────────────────┐
│ Frontend — Next.js 15 (App Router, TypeScript, Tailwind)     │
│  Landing → AI Trading cockpit → Wallet → Business (collapsed)│
└──────────────────────────────────────────────────────────────┘
        │
        ▼
┌──────────────────────────────────────────────────────────────┐
│ API Routes (app/api)                                         │
│  /api/trading/overview  — tick-on-demand cockpit state       │
│  /api/trading/actions   — pause / resume / stop / reset      │
│  /api/trading/live/*    — agent wallet lifecycle (SIWE-auth) │
└──────────────────────────────────────────────────────────────┘
        │
        ▼
┌──────────────────────────────────────────────────────────────┐
│ Trading engine (lib/trading)                                 │
│  hyperliquid.ts — market data client (meta/ctxs/candles)     │
│  strategies.ts  — momentum + funding signals (pure)          │
│  risk.ts        — sizing, TP/SL, circuit breakers            │
│  agent.ts       — scan → signal → risk → execute → report    │
│  exchange.ts    — signed live orders (EIP-712)               │
│  keys.ts        — per-user agent key custody (AES-256-GCM)   │
│  store.ts       — paper state (file storage, tmp fallback)  │
└──────────────────────────────────────────────────────────────┘
        │
        ▼
  Hyperliquid (public info API + signed exchange API)

Quick Start

# Install
pnpm install

# Configure (see .env.example / ENV_SETUP.md)
cp .env.example .env.local

# Develop
pnpm dev                # http://localhost:3000

# Run the trading agent headlessly (paper mode)
pnpm trading:agent

# MCP server — control the agent from Claude / any MCP host
pnpm mcp:stdio           # stdio transport (see docs/MCP.md)
pnpm mcp:test            # smoke-test tools/list + tools/call

Testing & Quality Gates

CommandWhat it runs
pnpm testJest — 1035 unit/property tests
pnpm test:uiPlaywright — browser E2E (own dev server on :3100)
pnpm lintESLint 9 (0-error gate; legacy debt as warnings)
npx tsc --noEmitTypeScript typecheck
pnpm buildProduction build (Vercel deploy gate)
go test ./...Go microservices (in services/*)

CI runs the full frontend suite (tsc + jest + lint) and Go tests on every push — see .github/workflows/ci.yml.

Documentation

DocumentDescription
PRD v2 — AI TradingProduct spec: positioning, cockpit, wallet architecture, strategy & risk parameters
MCP ServerControl the trading agent from Claude Desktop / Claude Code / any MCP host
Usage GuideIn-product guide (getting started, cockpit, going live, FAQ)
Risk DisclosureRead before live trading
Threat ModelAssets, trust boundaries and known limitations (uncertain orders, ledger drift, key custody)
Terms / PrivacyLegal
ENV_SETUP.mdEnvironment configuration
TESTING_GUIDE.mdTest infrastructure details
Agent APIProgrammatic surface for agents: policy, endpoints, errors, billing
Agent API contractOpenAPI 3.1 for the five agent endpoints — policy schema, error classes, real examples
Agent API casesFour worked integrations with real requests and responses
Agent toolsOne tool manifest for the Vercel AI SDK, LangChain, OpenAI and Anthropic
Framework startersRunnable Vercel AI SDK, Mastra and elizaOS integrations
PartnersPartner integration guide (curl, SDK, sandbox, billing)

Legacy payment-era docs remain in docs/ for the Business feature set.

Security Model

  • SIWE + JWT authentication for wallet users (lib/auth/siwe.ts, lib/middleware/api-auth.ts)
  • Non-custodial: user keys never leave their wallet; agent keys are trading-only and encrypted at rest
  • Risk guardrails enforced before every order (lib/trading/risk.ts)
  • Per-user isolation: TradingAccount / TradeRecord models with wallet-address scoping
  • Rate limiting, CSRF, replay protection, security headers (middleware.ts)

License

GNU General Public License v3.0 only (GPL-3.0-only) — see LICENSE and ADDITIONAL_LEGAL_TERMS.md.


中文文档

项目简介

Protocol Bank 是一款 AI 自动交易产品。用户连接钱包(MetaMask),在 Hyperliquid 存入交易预算,批准一个仅有交易权限的 agent 钱包后,AI 代理全天候交易永续合约市场——配有严格风控、全程透明、随时可撤销。Paper 模式使用真实行情 + 模拟资金,任何人都可以零风险体验。

核心安全属性:

  • AI 只能交易,永远无法提现——agent 钱包在 Hyperliquid 上仅有交易权限
  • 最大亏损永远写在屏幕上——交易钱包余额即最大可亏损额
  • 风控内建于每笔订单——止损、仓位上限、日内亏损熔断
  • 随时可撤销——暂停、紧急停止、或在 Hyperliquid 上撤销 agent
核心功能
模块说明
AI 交易驾驶舱 (/trading)登录后落地页:余额、净值曲线、自然语言 AI 活动流、持仓、暂停/停止控制
Paper 引擎真实 Hyperliquid 行情 + 模拟成交。动量(24h z-score)+ 资金费率信号,±2.5% 止盈止损,15% 仓位,5%/8% 日内熔断
Live 模式Agent 钱包批准流(EIP-712 approveAgent)、AES-256-GCM 密钥托管、由 agent 密钥签名的真实 IOC 下单、不确定订单安全停机、人工审批模式(可选)、每用户隔离 schema
钱包 (/balances)多链余额与流水(合并自旧 Dashboard)
Business(折叠区)企业支付套件:批量支付、发票、订阅、收单、联系人——保留但不再主推
快速开始
pnpm install
cp .env.example .env.local   # 见 ENV_SETUP.md
pnpm dev                     # http://localhost:3000
pnpm trading:agent           # 后台常驻运行交易 agent(paper 模式)
pnpm mcp:stdio               # MCP server:在 Claude 里查询/控制 agent(见 docs/MCP.md)
测试与质量门禁
命令内容
pnpm testJest — 1035 项单元/属性测试
pnpm test:uiPlaywright 浏览器 E2E
pnpm lintESLint 9(0 错误门禁)
npx tsc --noEmit类型检查
pnpm build生产构建(Vercel 部署门禁)
go test ./...Go 微服务测试
文档
文档说明
PRD v2 — AI 交易产品规格
MCP Server在 Claude 等 MCP 客户端里查询/控制交易 agent
使用指南上手、驾驶舱、上线流程、FAQ
风险披露实盘前必读
条款 / 隐私法律文档

⚠️ 风险提示:自动交易可能导致交易钱包全部损失。AI 是工具而非收益保证,历史表现不代表未来结果。Protocol Bank 不提供投资建议。

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.