Bitget MCP Server logo

Bitget MCP Server

Organization
Bitget-AI

Official Bitget Agent Hub — open-source AI toolkit connecting Claude, Cursor, Codex & ChatGPT to Bitget crypto trading (89 UTA v3 ops) and market analysis. Central installer for CLI, MCP, SDK, skills & signals.

PublisherBitget-AI
Repositoryagent_hub
LanguageJavaScript
Forks
21
Stars
228
Available tools
0
Transport typestdio
Categories
LicenseMIT
Links
  • Connect tools to AI workflows

    Bitget MCP Server exposes MCP capabilities that can be used by compatible AI clients and agents.

  • 0 available tools

    Browse the callable actions below, including names and descriptions when provided by the server.

  • Ready-to-copy setup

    Use the installation snippets to configure this server in your preferred MCP client.

  • Open source signals

    228 stars and 21 forks from the linked repository.


Overview

Bitget Agent Hub is Bitget's official AI Agent toolkit, covering cryptocurrency and tokenized US stock trading — bringing 89 Bitget UTA v3 operations across trading and market analysis to Claude, Cursor, Codex, ChatGPT, and other AI agents.

Built for individual traders, quant developers, and anyone who wants to operate a crypto exchange through natural language.

This repository is the central entry point for the entire ecosystem. See individual package repos for detailed documentation.


Quick Start

What Do You Want to Do?

What do you want to do with AI?
|
+-- Let an AI Agent trade on Bitget for me
|     |
|     +-- Terminal AI (Claude Code / Codex CLI / OpenClaw)
|     |   -> install agent-cli + agent-skill
|     |
|     +-- Desktop AI (Claude Desktop / Cursor / Continue)
|         -> install agent-mcp
|
+-- Let an AI Agent monitor and analyze the market for me
|     -> install bitget-signal (no account, no API key)
|
+-- I'm a developer building my own Bitget AI tool
      -> install bitget-agent-sdk

Install with Your AI

The use-case tree above tells you which package you need. To set up the terminal tools in one shot, paste this into your terminal AI (Claude Code, Codex, OpenClaw):

prompt
Please run `npx @bitget-ai/bitget-agent-installer upgrade-all --target all` to set up the Bitget Agent Hub terminal tools (requires Node.js 20+), then help me verify the install.

This installs the bgc CLI plus the trading and market-signal skills, and deploys the skills to Claude Code, Codex, and OpenClaw. Desktop AI (Cursor, Claude Desktop, ChatGPT) uses bitget-agent-mcp instead — see Trading Tools. For per-package commands, rollback, and flags, see Installer below.


Ecosystem

Bitget Agent Hub consists of 5 core packages, each with a single responsibility. This repo also ships the bitget-agent-installer meta-tool that installs and upgrades them together.

Trading Tools (API Key Required)

PackageBest ForWhat It Does
bitget-agent-sdkDevelopersFoundation TypeScript SDK with 89 UTA v3 operations
bitget-agent-cliTerminal AIbgc command-line tool for Claude Code, Codex, OpenClaw
bitget-agent-mcpDesktop AIMCP server for Claude Desktop, Cursor, Windsurf, ChatGPT
bitget-agent-skillTerminal AIAI reasoning guide teaching how to use bgc correctly

Market Tools (No Account Needed)

PackageWhat It Does
bitget-signal5 market analysis skills — macro, on-chain, sentiment, technical, news

All packages are MIT-licensed and free to use. They can be installed side by side without conflicts.


Installer

@bitget-ai/bitget-agent-installer is the ecosystem's meta-tool. One command installs, upgrades, rolls back, and deploys the Bitget Agent Hub packages into your AI tools — no installing each package and running each skill's setup by hand.

bash
npx @bitget-ai/bitget-agent-installer

Run it with no arguments for an interactive menu, or pass a command directly (see below).

What It Manages

PackageRole
bitget-agent-cliThe bgc terminal trading CLI
bitget-agent-skillTrading reasoning guide for terminal AI
bitget-signalMarket-analysis skills (no API key)

Desktop AI uses bitget-agent-mcp, which is registered in your MCP client's config rather than installed globally — set it up via the MCP package directly. bitget-agent-sdk is a library dependency and needs no separate install.

Why Use It

  • One command for the whole ecosystem — install, upgrade, or roll back every package together.
  • Always up to date — resolves each package's latest published version automatically; nothing to pin, so differing versions across packages just work.
  • Skill deployment built in — wires skills straight into Claude Code, Codex, and OpenClaw via --target.
  • Reversible — roll any package back to a previously published version.
  • Safe to preview — --dry-run prints every command without running it.
  • Lightweight — a single dependency-free script run with npx; auto-detects npm or pnpm.

Commands

CommandWhat It Does
(no arguments)Interactive menu — upgrade, rollback, or deploy skills
upgrade-all [--target <tools>]Install or update all managed packages to latest
upgrade <pkg> [--target <tools>]Install or update one package to latest
rollback <pkg> --to <version>Switch a package to a specific published version
install [pkg] [--target <tools>]Deploy an already-installed package's skills to your AI tools

Flags

  • --target <tools> — where to deploy skills: claude, codex, openclaw, all (comma-separated; default claude)
  • --dry-run — preview commands without executing them
  • --version · --help

Examples

bash
# Install everything and deploy skills to every supported AI tool
npx @bitget-ai/bitget-agent-installer upgrade-all --target all

# Update just the CLI to the latest version
npx @bitget-ai/bitget-agent-installer upgrade @bitget-ai/bitget-agent-cli

# Roll the CLI back to the pre-v3 release
npx @bitget-ai/bitget-agent-installer rollback @bitget-ai/bitget-agent-cli --to 1.2.0

# Deploy already-installed market-signal skills into Claude Code and Codex
npx @bitget-ai/bitget-agent-installer install @bitget-ai/bitget-signal --target claude,codex

# Preview an upgrade without changing anything
npx @bitget-ai/bitget-agent-installer upgrade-all --target all --dry-run

Trading Tools

Let an AI Agent Trade Bitget for You Automatically

Bitget Agent Hub's trading tools expose 89 UTA v3 operations through 14 curated intent verbs:

Ask Your AIWhat It Does
"Buy 0.1 BTC at market price"Spot market order
"Open a BTC long, 10x leverage"Futures position
"Check my balance"Account overview
"Transfer USDT to futures"Internal fund transfer
"Cancel all open orders"Batch cancel

Choose Your AI Environment

Your AI Lives In…Package Installed
Terminal AI: Claude Code, Codex CLI, OpenClawbitget-agent-cli + bitget-agent-skill
Desktop AI: Claude Desktop, Cursor, Continue, Windsurf, ChatGPT Desktopbitget-agent-mcp

Get Started

  1. Create a Bitget API Key at bitget.com/api-management
  2. Run the installer wizard or follow individual package docs
  3. Verify installation by asking your AI: "What Bitget tools are available?"

First time? Create a Demo API Key and use --paper-trading to practice before touching real funds.


Market Tools

Let an AI Agent Monitor and Analyze the Market in Real Time

bitget-signal adds 5 market analysis skills — no Bitget account, no API key required:

Ask Your AISkill
"How is Fed policy affecting BTC?"macro-analyst
"Are whales moving assets on-chain?"market-intel
"Is the market fearful or greedy?"sentiment-analyst
"Is BTC RSI overbought?"technical-analysis
"What's important in crypto today?"news-briefing

Get Started

bash
npx @bitget-ai/bitget-signal --target all

Restart your AI tool, then ask about market conditions.

Coming Next

Bitget-exclusive signals:

  • top-trader-flow: Top copy-trading leader positioning
  • derivatives-structure: Perp basis and term structure
  • large-flow-detect: Whale-sized order detection

Architecture

mermaid
graph TD
    subgraph Surfaces
        A[bitget-agent-cli<br/>Terminal AI]
        B[bitget-agent-mcp<br/>Desktop AI MCP]
        C[bitget-agent-skill<br/>AI Reasoning Guide]
    end
    
    subgraph Foundation
        D[bitget-agent-sdk<br/>TypeScript SDK]
    end
    
    subgraph MarketAnalysis
        E[bitget-signal<br/>5 Analysis Skills]
    end
    
    A --> D
    B --> D
    C --> A
    D --> F[Bitget UTA v3 API]
    E --> G[Public Market-Data MCP<br/>no API key]
    
    style D fill:#f9f,stroke:#333,stroke-width:2px
    style F fill:#bbf,stroke:#333,stroke-width:2px

How it works:

  1. SDK (bitget-agent-sdk): Foundation layer with 89 operations, HMAC signing, rate limiting
  2. CLI (bitget-agent-cli): Terminal command-line tool built on SDK
  3. MCP (bitget-agent-mcp): Desktop AI server built on SDK
  4. Skill (bitget-agent-skill): Teaches terminal AI how to use CLI correctly
  5. Signal (bitget-signal): Independent market analysis via a public, no-key market-data MCP service

Supported Operations

Based on Bitget UTA v3 API — 89 operations via 14 curated intent verbs:

ModuleDefaultIntent VerbsRequires API Key
market✅marketNo (public data)
trade✅order, position, strategy_orderYes
account✅account_overview, transfer_funds, …Yes
strategy—via rawYes
cryptoloans—loanYes
tax—taxYes

strategy_order is gated on the trade module — placing strategy orders is a trading capability an agent expects whenever trading is enabled; the strategy module's other operations remain reachable through raw.

Plus meta tools: discover (explore operations) and raw (call any operation by operationId).

Browse the live catalog anytime with bgc discover (or discover in any surface).


Security

Credential Protection

  • ✅ Never leaves your machine: API keys read from environment variables only
  • ✅ Local signing: All authenticated requests signed with HMAC-SHA256 in-process
  • ✅ No .env parsing: Tools do not read .env files to prevent leakage
  • ✅ No proxy: Direct connection to Bitget's official API

Safety Modes

  • --read-only: Removes all write tools at startup — AI cannot place orders or transfers
  • --paper-trading: Routes to Bitget's Demo environment — no real funds involved
  • Rate limiting: Built-in client-side throttling prevents API abuse

Best Practices

  1. Use Demo API Key first for testing
  2. Start with --read-only mode to verify queries
  3. Never commit credentials to version control
  4. Rotate API keys every 90 days

Documentation

TopicDoc
Getting starteddocs/getting-started.md
Ecosystem architecturedocs/architecture.md

Per-package references — CLI flags, MCP client config, SDK API, signal skills — live in each package's own repo (see the Ecosystem table above). The live tool surface is always browsable with bgc discover (or discover in any surface).


Links


FAQ

What is Bitget Agent Hub?

Bitget Agent Hub is the official open-source AI Agent ecosystem from Bitget. It connects Bitget's cryptocurrency trading capabilities and real-time market data to AI agents like Claude, Cursor, Codex, and ChatGPT — covering 89 UTA v3 operations through 14 curated intent verbs, plus no-account-required market analysis skills.

How do I know which package to install?

It depends on your AI tool:

  • Terminal AI (Claude Code, Codex) → bitget-agent-cli + bitget-agent-skill
  • Desktop AI (Claude Desktop, Cursor) → bitget-agent-mcp
  • Market analysis only → bitget-signal (no API key needed)
  • Developer building custom tools → bitget-agent-sdk

Can I use AI to trade crypto automatically?

Yes. Once installed, you can tell supported AI tools to "buy 0.1 BTC" or "open a BTC long at 10x leverage" and the AI will execute directly on your Bitget account. Use --paper-trading mode to rehearse safely.

Is it safe to connect my API key?

Yes. Credentials are:

  • Read from environment variables only
  • Never proxied through any server
  • Signed locally with HMAC-SHA256
  • Never logged or written to disk

How do I prevent accidental orders?

Add --read-only at startup. All write tools are removed entirely — the AI cannot see or call them.

Which AI tools are supported?

Claude Desktop, Claude Code, Cursor, Codex CLI, Continue, Windsurf, ChatGPT Desktop, OpenClaw, and any MCP-compatible client.

Is it free?

All packages are MIT-licensed and free to use. bitget-signal is maintained by Bitget and free to use — no API key, no account.

How do I update everything?

Run npx @bitget-ai/bitget-agent-installer and choose upgrade, or:

bash
npx @bitget-ai/bitget-agent-installer upgrade-all --target all

Contributing

Issues and pull requests are welcome.


License

MIT License — Free for personal and commercial use. Attribution appreciated but not required.


⚠️ Risk Disclaimer: Trading cryptocurrency and tokenized stocks carries substantial risk. You are solely responsible for any orders your AI agent places on your behalf. Always use --read-only and --paper-trading modes to rehearse safely before going live.

Bitget Agent Hub is an official Bitget open-source project · Part of the Bitget AI ecosystem · Built for traders, developers, and AI enthusiasts

Use Bitget MCP Server MCP with multiple AI models

TypingMind connects MCP tools at the workspace level, so once Bitget MCP Server is connected, you can use it with different AI models in TypingMind instead of setting it up separately for each model. This MCP runs locally through the TypingMind MCP connector on your device.

Setup guide to use the local connector

Use this when the MCP server needs access to local files, apps, or private resources on your computer.

1

Open the MCP settings

In TypingMind, go to Settings, Advanced Settings, then Model Context Protocol and choose Setup Connector.

  1. Open TypingMind in your browser.
  2. Click the Settings icon.
  3. Go to Advanced Settings.
  4. Open the Model Context Protocol section.
  5. Click Setup Connector and choose This Device.
TypingMind MCP connector setup screen with This Device selected
2

Run the connector command

Choose This Device, copy the command from TypingMind, and run it in Terminal. Keep the process running while you use MCP.

  1. Copy the setup command shown by TypingMind.
  2. Open Terminal on macOS or Windows Terminal on Windows.
  3. Paste and run the command.
  4. Approve the package install if Terminal asks you to proceed.
  5. Keep the Terminal window running while using MCP tools.
3

Add Bitget MCP Server as a server

When the connector status is Ready, click Edit Servers and paste the MCP server configuration.

  1. Wait until the connector status shows Ready.
  2. Click Edit Servers.
  3. Paste the Bitget MCP Server MCP server configuration.
  4. Save the server list.
  5. Refresh if you want to confirm the connector is still ready.
TypingMind MCP settings showing active server and Edit Servers button
{
  "mcpServers": {
    "agent-hub": {
      "command": "npx",
      "args": [
        "-y",
        "<mcp-server-package>"
      ]
    }
  }
}
4

Use it across models

Save the server list, open Plugins, enable the Bitget MCP Server MCP tools, then select any supported AI model in TypingMind and use the tools in chat or assign them to an AI agent.

  1. Open the Plugins page in TypingMind.
  2. Enable the Bitget MCP Server MCP tools.
  3. Start a chat and choose the AI model you want to use.
  4. Use the MCP tools in chat or assign them to an AI agent.
  5. Switch to another AI model whenever needed without reconnecting MCP.
TypingMind chat using enabled MCP tools with a selected AI model
Can you use Bitget MCP Server to help me with this task?
Bitget MCP Server
Sure. I read it.
Here is what I found using Bitget MCP Server.

Frequently asked questions

What is the Bitget MCP Server MCP server used for?

Bitget MCP Server is an MCP server that lets compatible AI clients connect to external tools and context. In TypingMind, you can add this MCP server once and make its tools available in your AI workspace.

Can I use Bitget MCP Server MCP with multiple AI models in TypingMind?

Yes. TypingMind connects MCP tools at the workspace level, so you can use Bitget MCP Server with different AI models such as Claude, ChatGPT, Gemini, or other models you have configured in TypingMind without setting up the MCP server separately for each model.

Why use Bitget MCP Server MCP with TypingMind?

TypingMind is one of the best frontends for LLM chat because it brings multiple AI models, prompts, plugins, AI agents, API keys, and MCP tools into one workspace. With Bitget MCP Server connected, you can use its MCP tools across your preferred models while keeping your chat workflow organized in TypingMind.

How do I connect Bitget MCP Server MCP to TypingMind?

Bitget MCP Server runs through the TypingMind local MCP connector. This is best when the MCP server needs access to local files, desktop apps, command-line tools, or private resources on your computer.

What tools does Bitget MCP Server MCP provide in TypingMind?

Bitget MCP Server exposes MCP capabilities that can be enabled from the TypingMind Plugins page and used in chat or assigned to AI agents.

Do I need to share my API keys with TypingMind to use Bitget MCP Server MCP?

No. TypingMind is local-first and lets you keep your model providers, API keys, prompts, and MCP configuration under your control. If Bitget MCP Server requires authentication, add the required headers, OAuth settings, or local configuration for that MCP server when you create the connection.

Related MCP Servers

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇