upbit
MCP Server for the Upbit API, designed to be a trading assistant specialized in technical analysis
claude mcp add --transport stdio scalalang2-upbit-mcp-server Path to mcp server \ --env UPBIT_ACCESS_KEY="Your upbit access key" \ --env UPBIT_SECRET_KEY="Your upbit secret key"
How to use
This MCP server provides a set of Upbit exchange tools exposed via the MCP protocol. It includes account management and trading capabilities (GetAccounts, PlaceBuyOrder, PlaceSellOrder, CancelOrder, GetAvailableOrderInfo, GetClosedOrderHistory, GetOpenOrderList), market data retrieval (GetMarketSummary, GetMarketTrends, GetDayCandles, GetWeekCandles, GetMonthCandles, GetMinuteCandles), and technical indicators (GetMovingAverage, GetMACD, GetBollingerBands, GetRSI, GetOBV). You can connect to this server using your MCP runtime and request these commands to interact with your Upbit account, fetch live market data, or compute common indicators for strategy development. The server requires your Upbit API keys to authenticate trading-related requests, while market data and indicators can be accessed with proper authorizations as defined by your MCP setup.
To use the tools, first configure your MCP client to point at the Upbit server with the provided environment variables (keys). Then send command requests like GetMarketSummary for a specific market, or PlaceBuyOrder/PlaceSellOrder to execute trades. For indicators, issue the corresponding GetMovingAverage, GetRSI, and other indicator requests with the target symbol and desired parameters (window size, period, etc.). The server aims to provide a cohesive set of Upbit capabilities through MCP without requiring direct client integration with the Upbit API.
How to install
Prerequisites:
- A running MCP runtime (Node, Python, or your MCP host) capable of loading a server defined under mcpServers.
- Access to the repository containing this Upbit MCP server.
- Upbit API credentials to use for trading actions (if you intend to place orders).
Step 1: Clone the repository
git clone https://github.com/your-org/scalalang2-upbit-mcp-server.git
cd scalalang2-upbit-mcp-server
Step 2: Install dependencies (choose the appropriate path for your runtime)
- If this server is provided as a Node.js project:
npm install
- If this server is provided as a Python project:
pip install -r requirements.txt
Step 3: Configure environment variables
- Create a configuration file or export environment variables in your runtime to provide Upbit credentials.
export UPBIT_ACCESS_KEY=your_access_key
export UPBIT_SECRET_KEY=your_secret_key
Step 4: Start the MCP server using your runtime
- If using a Node-based MCP host:
npx <your-mcp-host> start --config path/to/config.json
- If using a Python-based runtime (uv):
uvicorn main:app --reload
Step 5: Verify the server is running
curl http://localhost:PORT/health
Note: Replace placeholders with your actual package names and exposure endpoints as per your MCP runtime documentation. The key point is to provide the Upbit API keys via the env block or equivalent in your runtime configuration.
Additional notes
Tips and considerations:
- Keep your Upbit keys secure; use scoped keys and rotate them periodically. Do not commit keys to version control.
- Ensure proper error handling for trading actions (PlaceBuyOrder, PlaceSellOrder, CancelOrder) to handle insufficient balance, invalid market symbols, or order rejections from Upbit.
- For market data and indicators, consider caching frequent queries (e.g., GetMarketSummary, GetMinuteCandles) to reduce API calls and latency.
- If running multiple servers or accounts, namespace requests by the server name (e.g., Upbit) to avoid cross-env conflicts.
- Check rate limits in your MCP client and Upbit API limits to prevent throttling.
- Enable logging for easier troubleshooting, including request payloads, responses, and error traces.
Related MCP Servers
trpc-agent-go
trpc-agent-go is a powerful Go framework for building intelligent agent systems using large language models (LLMs) and tools.
station
Station is our open-source runtime that lets teams deploy agents on their own infrastructure with full control.
tiger-cli
Tiger CLI is the command-line interface for Tiger Cloud. It includes an MCP server for helping coding agents write production-level Postgres code.
gopls
MCP server for golang projects development: Expand AI Code Agent ability boundary to have a semantic understanding and determinisic information for golang projects.
kubernetes
A Model Context Protocol (MCP) server for the Kubernetes API.
gcp-cost
💰 An MCP server that enables AI assistants to estimate Google Cloud costs, powered by Cloud Billing Catalog API and built with Genkit for Go