ryckli/cryptoagentpro.beta — explained in plain English
Analysis updated 2026-05-18
Backtest a crypto trading strategy against a year of historical price data before risking money.
Run simulated paper trades to see how a strategy would have performed in real time.
Get AI generated suggestions every 15 minutes about whether to switch trading strategy.
Build a custom trading strategy from indicators like RSI or MACD without writing code.
| ryckli/cryptoagentpro.beta | agricidaniel/gogh | betta-tech/harness-sdd | |
|---|---|---|---|
| Stars | 46 | 46 | 46 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | moderate | easy |
| Complexity | 4/5 | 3/5 | 2/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Needs an exchange API key and a DeepSeek AI API key, plus optional MongoDB and Redis for full functionality.
CryptoAgentPro.beta is a cryptocurrency trading system built around an AI agent, based on the architecture of a project called TradingAgents-CN. The system combines six built in trading strategies plus support for custom ones, an AI model called DeepSeek that checks market trends every fifteen minutes, a backtesting engine that can replay history at speeds up to one hundred times normal, and two safe trading modes so nobody accidentally risks real money by mistake. The two trading modes are paper trading, which simulates trades locally and is the default safe option, and testnet trading, which sends orders to an exchange's sandbox environment where no real funds are at risk. Both modes can be switched with one click from the web frontend. Before any order goes through, a risk control gate checks it against limits such as a maximum loss of three percent of capital on a single trade and a maximum daily loss of fifteen percent, automatically halting trading if that threshold is hit. The AI component analyzes the coins a user is monitoring every fifteen minutes, produces a report, and suggests whether to switch trading strategies. By default a human must approve any suggested switch before it happens, though full automatic mode can be turned on. Users can also build their own custom strategies from indicators like EMA, RSI, MACD, and BOLL directly in the web interface, and adjust API keys, risk settings, and monitored coins there as well. Under the hood, the project is split into a core trading engine written in Python, a FastAPI backend that exposes the trading and analysis features as an API, and a Vue.js frontend styled with Tailwind CSS. It needs Python 3.11 or newer and Node.js 18 or newer to run, with MongoDB and Redis as optional add ons since it can fall back to a simpler SQLite database. A Docker setup is included for running the whole system together as a set of services.
An AI powered crypto trading system that runs six trading strategies, checks market trends every 15 minutes, and lets you test safely before risking real money.
Mainly Python. The stack also includes Python, FastAPI, Vue.js.
Setup difficulty is rated moderate, with roughly 1h+ to a first successful run.
Mainly developer.
This repo across BitVibe Labs
double-check against the repo, no cap.